About WBL, what it does or doesn’t do for you
WBL is a measure of expectation, the expectations the designers had for items and wealth. These expectations affect other aspects of the game—perhaps most notably monster design. Monster save DCs account for cloaks of resistance, their own save bonuses account for the ability-score-enhancing headbands, and so on. To a certain degree, the CR of a monster is based on the assumption that certain tools will, or won’t, be available.
This is important. I would caution very strongly against mucking with it too much, and would suggest that the game does not work well if you do unless you also put a lot of effort into addressing how things will work with more or fewer wealth. I also routinely remind people that it’s crucial to remember that this is wealth we are talking about—the sum total value of the current gear enjoyed by the character. Consumables already used, obsolete equipment that’s been sold off, that isn’t part of wealth. And wealth is best calculated using its value to the character in question, which may be different from what the book says it is or what NPCs are willing to pay for it. In other words, wealth should only count if it’s actually usable, because WBL is supposed to be a measure of power derived from property. That is the sole reason it has any use.
At the same time, WBL can only, at best, allow you to match designers’ expectations. This can allow CR and other tools to work more reliably. But these tools aren’t all that reliable, even at the best of times. This means that keeping to WBL is very far from a magic bullet—but it also means that straying too far makes a bad situation worse, so take that for what you will. And something really important: this goes in both directions. In fact, going under WBL is often worse than going over WBL, assuming the magnitude of the difference is equal.
Moreover, there are concerns aside from balance to worry about—narrative flow, consequences for player actions, and so on—that also have to be accounted for. For instance, selling off obsolete items only gets you half that item’s value. Used consumables don’t mean you immediately stumble upon a tidy little pile of gold pieces equal to its value. And gold, for that matter, doesn’t immediately or automatically turn into the items you want, and money in your pocket doesn’t actually help you with most challenges in that form. Because it would be absurd if it did, and because it would mean that, e.g., consumables wouldn’t actually cost anything, which isn’t quite right.
What this all amounts to is that WBL is a rough goal, not a fixed state. Wealth should, and indeed must, fluctuate around WBL. This is absolutely required and certainly not a “problem” that needs immediate remedy. Attempting to do so is often heavy-handed and blatantly metagamey, which can and does ruin immersion at times. In severe situations, that is entirely appropriate, but since wealth is always fluctuating, doing so every time wealth strays from WBL is going to result in doing your game more harm than good.
Finally, one last note on what WBL is decidedly not—economics. WBL blatantly flies in the face of the very concept of economics, an arbitrary value expected without any mechanism whatsoever in place determining that value. That is because WBL is about character power far, far more than it is about money. It’s a purely game-design issue, not an economic one. And this is important because Pathfinder does not have sane or stable economic rules. They cannot withstand even the barest critical inspection; WBL basically states that gp has to go up in concert with XP, but almost all of the minimal economics in the game are largely divorced from XP-acquiring activity. This is generally “balanced” by the fact that most of them offer extremely small amounts of money relative to simply going out and acquiring loot, but not all wealth-generating activities are. WBL itself, for that matter all of Pathfinder, offers no solution to this dilemma. The general assumption is that it won’t matter because Pathfinder characters are “supposed” to go out and adventure. It’s another expectation baked into the system.
Now then, with an overly-lengthy introduction finished, let’s discuss remedies.
Before the game/out of game
System buy-in
As discussed above, Pathfinder largely ignores these problems because it’s assumed that players will largely do so as well, because their interest is in adventuring. After all, they are playing Pathfinder, that’s what the system is for. It’s based on Dungeons & Dragons, a system developed for and focused on dungeon-delving and dragon-slaying. To a certain extent, Pathfinder doesn’t have, and doesn’t want, sophisticated mechanisms for handling economic activity. It’s the wrong system for that.
When the group is playing Pathfinder, there should be at least the minimal buy in that the group is playing an adventure and their characters are adventurers. If that isn’t the case, if characters seriously want to stay in safe locations and participate in the world’s economy, then Pathfinder is a terrible system to use for the game. It doesn’t have any depth whatsoever in that area, and the rules it does have will hurt the campaign more than they will assist.
Bonus XP can be awarded for cleverly making money
This can be dangerous, because it might encourage trying to ignore Pathfinder’s focuses and strengths and trying to use it for a game that would really be better off in another system. However, Pathfinder is also very limited in terms of its economics, and it still may be the right system for a campaign about adventurers who still have some economic prowess—and in that sort of situation, you need answers.
One answer for that can be to award XP for successful wealth-generating schemes. This can cause problems with your plans (you’ll have to make them tougher, since you’re dealing with higher-level, higher-wealth characters than you [or the AP] were planning on), but increasing level increases the expected WBL—which means that the players’ greater wealth is more in line with their level. The idea here is to fight skew, that is, characters who are too strong in one way but not particularly strong in another, which makes it difficult to find appropriate challenges. Increasing level in this way can actually make things easier to balance, since XP and gp are enjoying a relationship closer to what was expected.
Houserule to maintain expectations
Some tricks are just too good, generate too much wealth. Houserules to nix them are heavy-handed, but these kinds of things should be discussed out of game, and comes back to the idea of buy-in from the players. “Congratulations, that’s a clever idea, but it’s also going to mess up the game we’re trying to play, so we’re going to have to adjust it so it’s less good.” Hopefully, they’ll accept that (buy-in), and hopefully, you can do it ahead of time, before they sink too much into trying this scheme.
Talking to your players and having good communication about expectations and plans is key here—as it is in the entire game. Allowing the player some benefit, or to enjoy the full benefit briefly, can also be decent approaches to mitigate disappointment here. Framing also helps: “Your trick was so good that we had to change the rules to keep things competitive” is one of the most impressive statements that society has for competitors. Pathfinder isn’t a competitive game, but the feel-good effect of that kind of framing is still valid.
You can also houserule things to make the economy react in ways that are a little more realistic, to prevent tricks from turning into the kind of run-away success that generates substantially-higher-than-expected income on a continuous basis.
In-game
Behind-the-scenes number fixing
Subtlety in handling WBL is to be strongly encouraged—since wealth is fluctuating constantly, you do not want it to be obvious that you are managing it at all. Otherwise it will be rather like railroading, and have the same problems.
Instead, acquisition of significant wealth should result in advantages—temporarily. Loss of significant wealth should likewise result in disadvantages—temporarily. And you should quietly adjust the wealth that the players are seeing in later loot to bring things closer to WBL. It’s OK if they’re over or under for a level, possibly even two. WBL grows pretty quickly with level, so even if players find a major windfall, if they then gain XP without gaining gp for a while, they’ll often pretty quickly reach a level where the wealth they have is entirely appropriate. Likewise, adventuring often produces substantial finds all in one place.
And in the meantime, you can try to tweak challenges to account for deviations in wealth. This is more difficult, and the rules have less guidance for you, but it also has the advantage of allowing those with more wealth—who take on greater challenges than they otherwise might at their level—to acquire XP faster, while those with less wealth—who are constrained from taking on everything they might at their level with normal wealth—from gaining more XP too quickly and pushing the disparity further.
This is the bread-and-butter of managing WBL. This is how you should usually handle disparities. And it’s OK if it takes quite a bit of time. But it doesn’t handle truly dire situations where the numbers are simply off by too much.
Way too little wealth: side quests and non-combat challenges
Lack of WBL is a huge problem in Pathfinder, in part because there are classes that barely manage as it is, and rely heavily on the items they can acquire to shore myriad weaknesses. One great way to handle this situation is with challenges that provide targeted wealth (things characters desperately need), and do so without requiring wealth. Most of the “hard requirements” on wealth are about combat, so challenges that rely less on combat—either involving relatively simple fights or none at all—are great opportunities to introduce this wealth. This works best if it’s at least partially player-directed. When your party is struggling, suggest that the characters could take some time to apply their skills and regroup. Ask for Diplomacy checks, Survival checks, Knowledge checks, and so on to find long-lost treasures, rich patrons who need things done, and the like. It’s a large, magical world. As long as they aren’t on a forced march of a quest, they can take some time to find some of that wealth they’ve been missing.
This is also, by the way, a fantastic opportunity for plot hooks. A team of adventurers is struggling, needs better gear? A shady anonymous benefactor might be able to supply them, but now the team doesn’t know who they’re working with, don’t know what favors might get called in. Or maybe there was a reason that relic was left, untouched, in that temple for so long—and maybe it means someone in the party is the chosen one, or maybe it means there was a subtle and dire curse they’ll only learn about later. And so on.
Way too much wealth: you could just leave it alone
It is possible for players to gain far too much wealth and have far too powerful toys available for their level, without actually requiring a response from you. So they take on tougher enemies than they should, they accomplish more than they otherwise would—you can always just throw more at them. This gives them more XP and allows them to reach the level where their wealth is appropriate sooner. It accelerates the game, which can be a problem, but isn’t always. Not always appropriate, but an option to consider.
Obviously, only works if the wealth was the result of a one-time thing. Much-too-high income is more problematic.
Way too much income, or too much wealth you can’t leave alone: Get heavy handed
You do have to get heavy handed sometimes. Other answers here give many options for doing this, which mostly revolve around stripping characters of their wealth. This works, but it can feel bad, and it doesn’t address the problem of too much income if that’s what you’re dealing with.
As a result, you will probably have to combine this with a little bit of letting the players enjoy it. That is, the above suggestion of just leaving it alone will have to apply at least somewhat. This mitigates the “feels-bad-ness” some (by giving the players at least some time being overpowered), and gives more time, which may increase their level (reducing the problem even if it doesn’t fix it) and allow you to set up something more subtle than blatantly taking their toys away.
The ideal case for handling things when they get really bad is to make their tricks obsolete or inapplicable, rather than outright taking them away. Ultimately, the biggest problem with being overpowered, whether from having too much wealth or otherwise, is that it becomes too easy to just apply the same solution to all problems. When you have a really good hammer, all the world still looks like nails even when you have other stuff. Taking away the nails is often a better approach than taking away the hammer.
But sometimes you have to. I urge you to limit how often you do, to try to avoid that situation in the first place, but it’s part of the game. When you can do it judiciously, rather than regularly, it can also improve the game, even better if you never had the problem in the first place—after all, a big set-back like that is a great story.
Really severe problems with too-powerful toys: address out of game
If your campaign literally cannot continue without removing some item or items from the players’ possession, and you do not have a good in-character way to do so, even a heavy-handed one, just own up to the problem. GMing is not easy, mistakes happen. Your group should understand that. Admit you made a mistake allowing them to get the item(s) and that you need to remove them or nerf them to keep the game workable. Discuss as a group how they’d like that to go. Some might prefer some deus ex machina in-character explanation over nothing, others might prefer that the item gets quietly ignored and forgotten, and still others might rather you just ret-con it out of the game. They may have ideas for nerfing it without completely removing it, and they might have ideas on how to justify that in-character.
And when you do this as a group, almost all of the problems go away. The idea is to find a solution that allows the game to get back on track. To make people satisfied with where the game is going. A lot of the problems of being heavy-handed go away in this case.
I've just finished running this module, with great success, so I can address a few of your concerns..
First, I challenge the premise that Dead in Thay needs to be a "fast-paced assault". Your players signed up for this adventure not because it was a "fast-paced assault", but because it was a behemoth of a dungeon. They knew full well it would be a long slog. So give your players what they expect, which is a vast and varied dungeon loaded with combat and interesting encounters. Don't stress too much about the pace not quite being fast enough.
But it can still be very easy for Dead in Thay to become drawn out and long. Some of the encounters can get repetitive ("Oh, another Thayan Apprentice? And a wight? We've had ten of those already."), and stopping to rest for long periods of time can break the verisimilitude and reduce the dramatic tension.
You can keep the drama and enhance the verisimiltude by treating the Doomvault as an active and dynamic complex which responds visibly to the players' actions and pushes its own goals against the world around it. You can help the player characters push through the many encounters by providing them with extra resources. And you can keep the game relatively fast-pased by ensuring that the players stay focussed on the end goal and directing them towards the end.
A dynamic dungeon
The Doomvault is an active military complex. The Thayans within have a whole army and facilities to create monsters to bolster their ranks. The Thayans within are at war with the Thayan rebels keeping them inside. Do not think that the Thayans will just sit by idly as the PCs march in and start hacking and slashing their way through the Doomvault. These suggestions delay burn-out by making the Doomvault change over time.
Alert Level and scaling encounters: A simple change to make is to the Alert Level, which is a mechanic in Dead in Thay which makes the random encounters more dangerous the longer the players spend in the Doomvault. I have found that having the Alert Level go up by 1 per hour (rather than 1 per 4 hours) gives a nice progression. For the first day or two they had basic random encounters, whereas towards the end of the adventure they were getting higher tier random encounters. The earlier encounters might be flavoured as ordinary patrols, whereas later encounters may be squads sent specifically to hunt down the party. The scaling severity of these encounters will enhance the sense of danger the players feel the longer they spend in the dungeon.
PCs become (in)famous: The Doomvault is also occupied by intelligent NPCs. In the first day, the Thayans at large might not notice the PCs unless they make a big fuss, but by the second day word will have spread that there is a party of adventurers raiding the Doomvault. If some Thayan-aligned NPCs witnessed the PCs and lived to tell the tale, the Thayans will be well-informed about the party's composition and may start preparing countermeasures. If the PCs ensured there were no witnesses, then the Thayans will only know that someone has been slaughtering parts of their dungeon and that they need to be alert. Over time the PCs will become quite well known (and likely deeply feared) by the residents of the Doomvault. Your players will get a kick out of overhearing a conversation about their exploits or from an enemy who is cowed by their fierce reputation, and it makes them feel like their actions have impact.
Use the dungeon's facilities until they break: Part of the goal of the PCs is to disrupt activities in the Doomvault, so keep track of which activities the PCs disrupt, and which ones they don't disrupt. For example, before the party clears out the Ooze Grottos, you can send oozes to reinforce other parts of the dungeon. Once they kill most of the people in the Ooze Grottos, though, the Doomvault no longer has the capacity to send oozes to reinforce other parts of the dungeon. There are other examples. Two of my favourite facilities were the Undying Laboratory and the Golem Laboratories. While the Undying Laboratory still stood, I had it convert some of the corpses of monsters my players killed into new undead creatures, such as skeletal trolls or zombie gorgons, which presented some unique enemies for the players to fight. As for the Golem Laboratories, having the threat of golems looming over the players' heads was a great motivator for them, even if they fought only a few flesh golems. Note that if the players clear out an area, the dungeon loses the ability to send this sort of enemy. If my players had braved the Golem Laboratories and prevailed, they would have been safe from golems from that point onwards.
The dungeon regroups while the PCs rest: This is explicitly stated in the adventure module, although you might initially be at a loss as to how to make the dungeon recover without sending the PCs back to square one. Whenever the PCs take a long rest, you can replace some of the non-unique creatures in a room. Unique and high-level creatures, however, should stay dead. Consider the functioning of that room and what staff are needed to provide the minimum functionality. Also consider what the Thayans have to replace lost manpower. They have a sizeable number of humanoids, but that number is still finite. However, they also have several necromancers, so any slain humanoids can be raised as skeletons, zombies, wights, deathlock wights or dread warriors, providing ongoing supplies of those creatures. Additionally, the Doomvault can create oozes and golems and summon demons and aberrations, so these are also renewable sources of minions. Furthermore, the Thayans may reinforce strategic locations once they learn of the party's actions (golems are a favourite of mine), encouraging the players to push through and clear out an entire sector without taking too many rests.
Make resting unsafe, and the dungeon has its own goals: In the case of Dead in Thay, it is stated that the goal of the Thayans is to survive and to kill the party. However, I think this overlooks another goal, and that is to break the siege which the rebels have placed on the Doomvault. It would probably take a day tops for the Thayans to realise that Syranna has captured the gatehouse and is trapping them inside their dungeon. With the party slowly overwhelming their defences, the Thayans will want to break through to the gatehouse so they can deny the party their resting spot and get reinforcements from outside. After many sessions, I had the party walk into the gatehouse in time to witness the conclusion of a large battle between the rebels and the Doomvault Thayans. The rebels had won for now, but this event signalled clearly to the players that the gatehouse was not as safe as they once thought and that they needed to hurry up and finish the adventure. I planned for a series of progressively harder encounters set in the gatehouse for each time they long rested after that, although my party managed to finish the adventure before that. They pushed for a rapid conclusion because they felt that long resting was dangerous. This sort of tension adds greatly to the feeling of a "fast-paced assault". However, I caution that this tactic should not be used too close to the start of the adventure, but rather be used to hurry up the party when you get closer to the end.
Assist the party
The Doomvault is taxing on the party's resources, especially for small parties like your own. As you have discovered, you quickly blow through a whole adventuring day's worth of encounters in just a couple of in-game hours. The above points help you motivate the party to keep going despite the strain on their resources, but there will still come a point when the party simple doesn't have any resources left to keep fighting. As such, you need to help the party out.
Give them care packages: In Dead in Thay, the party is asked by the rebellion leader Syranna to raid the Doomvault. Syranna has a vested interest in the party's rapid success, so Syranna would leverage whatever resources she can to aid the party. To provide healing, I gave the party a liberal supply of healing potions. To deal with casters who were running out of spell slots, I gave them spell scrolls. If you want the party to go for longer without long resting, give them consumable resources which effectively grant extra uses of their limited use abilities or otherwise supplement them (e.g. spells, healing). Then I had Syranna tell the party not to come back until they had finished using those items.
NPC Allies: A simple remedy to a small party is to make the party larger. The Doomvault contains numerous NPCs who have the potential to become allies. Some of them might require a long rest to get back to full health, and some might need to be assigned a stat block, but they can be used to aid the party if they are under strength. You could also bring in NPC allies from previous adventures. Don't give them too many NPCs at once though, as that would slow down combat and take the spotlight away from the players. You can let the players keep some agency and take the pressure off you by letting the players determine the NPC allies' actions in combat (although you keep veto power to stop them from acting out of character). For Dead in Thay in particular, if the party lacks proficiency in Arcana, an NPC ally goes from useful to essential because many things in this adventure require Arcana checks.
Recruit more players: Ideally you want a large party for this adventure, not a small one. If at all possible, if you have a small party you can attempt to increase the size of your group by getting more players. This is not always possible (I wasn't able to do this), but if it is possible it is preferable to NPC party members.
Facilities for restoration: There are a couple of places in the Doomvault which provide benefits to characters which would otherwise require a rest. There is a magic pillar in one place which is an unlimited source of healing (with a small risk associated with it). There is also a pool in one place which grants the benefits of a short rest once per ten days, although that was woefully underpowered so I house-ruled it to be the benefits of a long rest (ignoring normal time limits) once per day. Such facilities allow the party to adventure for longer each day. You can guide the players to discover these facilities, and remind them of their existence whenever you think the adventure would benefit from their use.
Let them shop: There isn't much loot in the Doomvault, but there is some. In my campaign I had established the existence of a magic item merchant as a recurring NPC, so I had her show up during the party's extended long rests for them to trade if they wanted. This let them collect new assets which would help them in their adventures, and let them liquidate loot they had acquired from the Doomvault without having to wait for the end of this very long adventure. You will need to figure out something which makes sense for your campaign. In any case, your party will likely have many hours of downtime each day as they recover from adventuring, which can be spent shopping outside the Doomvault or crafting.
Goal-oriented raiding
In the massive Doomvault, it is very easy for the players to lose sight of why they are raiding this dungeon and just get stuck in the grind of mega-dungeon combat. But the party does have a goal, and it is your job to make sure the players stay on track.
Plentiful clues: One of the party's objectives is to determine the location of the Phylactery Vault. For this, you need to give out lore. The adventure module has a list of items of lore which you are meant to give in order and certain NPCs who can give a set number of pieces of lore. However, I recommend not sticking too strictly to this system. The items of lore are good, but some NPCs would know more than others, so you don't need to stick too closely to the order or the particular number of pieces of lore. The more clues you give out, the faster the players will be able to finish the adventure. I would lean towards being generous with sharing of lore.
Tell them to stay on track: In Dead in Thay, this can be done using the NPC Syranna, who the party answers to. Have Syranna encourage the PCs to keep going and not rest every 15 minutes. Have Syranna remind the party of what they need to look for. Have Syranna point out major leads the party should follow. Have Syranna ask the party regularly for their progress. Have Syranna intercept communications or track movements within the Doomvault to give the PCs hooks to go to certain parts of the Doomvault. Constant prodding and reminding will help keep the goal at the forefront of the players' minds.
Guide the party to more interesting areas: Unlike many dungeons, which have a fairly linear design, the Doomvault is almost a sandbox with 7 different entry points. However, some sectors are more interesting and plot-relevant than others. My party started by going into the Ooze Grottos then progressed to the Forests of Slaughter, which are two of the sectors with the fewest NPCs which can be interacted with outside of combat and next to no hooks to the greater goals. The players rather quickly got bored of these sectors and the game became a grind until they moved on. I would recommend providing the players with recommendations for good places to start or go which lead them away from the boring areas and into the interesting areas which have more plot hooks.
If you apply these points of advice, they should help keep the adventure from getting stale and push the players to progress through the adventure more quickly. While some of this advice is specific to Dead in Thay, it should be applicable to similar adventures too.
For reference, I'll give you a few of the details of my time running this adventure. For most of the adventure I had three players at each session, with a wizard, a paladin and either a druid or a barbarian. It took us 14 four-hour sessions to finish Dead in Thay, although your mileage may vary. The party took a total of 4 long rests throughout the adventure (although without the magic long-rest-granting pool, they would have taken 5 long rests). The party started at level 9, and by the end, before the final boss, the regular players had reached level 13 (although if I had a larger party they would have probably only reached level 12).
Best Answer
No, treasure is not low in Rappan Athuk
But that is due to the adventure's nature: It is a sandbox campaign.
What does that mean? There are treasures hidden in places you can't even imagine, but you have to look out for it. Not all treasure is in plain sight, rewarded after defeating enemies. Especially because you will hardly defeat all enemies in the campaign.
Sandbox campaigns in general have an issue where random encounters will not generate enough treasure, while at the same time inflating the XP (and thus the group's level), which makes it so you need even more treasure to be within the expected for your level. This may also happen if the adventure has a lot of traps (and it does), since those naturally won't have a lot of treasure (dead enemies are probably looted by denizens).
As a result, you will feel like the adventure is not rewarding you enough, and that even may be true for the first few levels, or based on the direction the adventure has moved. If your group happens to have picked the leads that won't award very well, it's just bad luck. If your group had a lot of outdoors encounters, either while travelling or while camping and got attacked by random enemies, then you likely won't have a lot of treasure either.
But keep in mind that it's ultimatelly a dungeon campaign, meaning that the focus of the designers was in the designing of the dungeons, and the deeper you go, more treasure you will be rewarded.
Crafting magic items does not shorten your wealth. It transfers your wealth from "unwanted treasures" to "wanted treasures". Example, if you dislike that an enemy dropped a +1 halberd because nobody in the group can use or wants to use it, then you can sell it for half price (market value 2,000 gp, thus 1,000 gp) and buy material components to make that +1 longsword that your fighter really could use (1,000 gp in components). See how the net wealth did not change? It is merely transferred from one type of item to another.
Have a little chat with your GM
See if he isn't ignoring the fact that some random encounters should have treasures with them, and he happened to forget about them because such items aren't always listed in enemies's statblocks. This doesn't have to be monetary or magic items, but are supposed to increase the party's wealth accordingly.
For instance, you know that Bonesuckers (and Vrocks) have a Standard treasure, right? Even if the monster can't really equip magic weapons and armors, their lair should have the remains of their previous victims, or you could rip their belly off and find that they had eaten a few precious gemstones, or even, you heard that there is a merchant paying a good sum of money for a few bonesucker tentacles.
So, it's possible that your GM forgot to account for this extra treasure from those encounters, and that's why your group is under the average wealth for your level.