There's no point to doing this. The idea behind spell points is to have a single resource fueling your spells regardless of their level. However, Warlocks already have that, because their spell slots are only of a single level.
With that said, if you insist on doing this, the correct progression would be:
- 2
- 4
- 6
- 6
- 10
- 10
- 12
- 12
- 14
- 14
- 21
- 21
- 21
- 21
- 21
- 21
- 28
- 28
- 28
- 28
This is a direct translation of the Warlock's spell slots at each level into spell points; which is exactly how spell points are calculated for all other classes.
Note that if you use this system and allow the Warlock to create lower level spell slots than their maximum, you are fundamentally changing the Warlock class. I would carefully consider balance before I did this. Of course, if you don't allow them to create lower level spell slots, there is absolutely no point using the spell point system, since it will just be a direct translation of their spell slots.
Based off of DDAL1 modules...
The Table:
\begin{array}{l c l}
\text{Total GP Expended} & \text{Level} & \text{(Expenditure for next level)} \\ \hline
0 & 1 & (700) \\
700 & 2 & (1,000) \\
1,700 & 3 & (1,350) \\
3,050 & 4 & (1,850) \\
4,900 & 5 & (2,500) \\
7,400 & 6 & (3,500) \\
10,900 & 7 & (4,750) \\
15,650 & 8 & (6,500) \\
22,150 & 9 & (8,750) \\
30,900 & 10 & (12,000) \\
42,900 & 11 & (16,500) \\
59,400 & 12 & (22,500) \\
81,900 & 13 & (31,000) \\
112,900 & 14 & (42,500) \\
155,400 & 15 & (58,000) \\
213,400 & 16 & \\
\end{array}
The Method:
I went ahead and tabulated the possible XP to earn and possible treasure haul2 for each of the two-dozen or so DDAL modules I've got on hand.
Each module is designed for one of the following level spans3: 1-2, 1-4, 5-10, or 11-16.
From the XP earned in each I calculated a fraction of the indicated level span that would be "traversed" by completing the module, and used that to extrapolate how much gold would accrue to one gaining a level in that span.
I then weighted each by the recommended hours of play4 and ran a power regression.5.
Finally there's just a bit of rounding to make the numbers... round.
The Application:
Full confession: I don't think this ^^ is the best way to come up with these numbers.6 But I think it's a way and figured it's worth letting voters see so that wiser heads than mine can decide.
Those numbers are just... insane. Dropping that much cash--per adventurer!--onto the population of Barovia is just ridiculous. I know that D&D doesn't try to model any sort of functioning economy, but this is a bridge too far for my credulity. So you've got to find ways to ameliorate it. Perhaps use this table as a party table, so that once PCs have dropped 700gp everyone bumps to level 2? Perhaps adapt Delta's advice and switch to a silver standard (for XP) so that this all drops by an order of magnitude? Or perhaps...
magical items should be allowed as part of this scheme. Buying everyone a round and carousing for a night doesn't feel (to me) very different from rescuing a villager from weres and tossing a potion of healing their way. But this could also be a place where you can exert a GM's thumb on the scale, since the "values" of any items that the players disposed of in a way that buffed their renown aren't terribly standard. (And to that end, I'd recommend the Sane Magic Item Prices index; it may not be perfect, but it's waaay better than following the DMG's loose guidelines.) And now you've given players the interesting choice between hanging onto their widget of frobbing or "cashing it in" to level up. (Or level everyone up, per point 2!)
1 - Dungeons & Dragons Adventurers League, WotC's organized play program that was (and sort-of is?) active during 5e. The modules published through this program followed a rough set of guidelines for XP and treasure given out, and I'm using this as an insight into what WotC employees (Chris Tulach, specifically, perhaps with input from other designers?) thought was a good pacing of treasure accumulation by level.
2 - only coinage; since you indicate your players won't really be able to "cash in" any magical items, we're going strictly off of currency here. But more on this later.
3 - I'm using the term "span" rather than "tier" as "tier" is a defined term in 5e (PHB p.15) and these modules don't all correspond to tiers. It feels clunky, but I didn't want to conflate the ideas.
4 - Some modules--only in the lowest two spans--are recommended for two hours, most are recommended for four. This then has the effect of dialing back the impact of these smaller modules. However, there are many more of them (fifteen in 1-2 and 1-4) than in the higher levels (six in 5-10, one in 11-16) so the model's still drawing more of its info from those levels.
This weighting toward low-level information strikes me as fine in one way, as that's where the majority of play tends to happen. On the other hand, this means that any "errors" in the model are likely to appear at the high end, where it's going to take a while for you to notice and be really hard for you to fix: "oh, dear, we've been on level 10 for five sessions now, and are only halfway through. And I want them at 12 to face Strahd et al. for the last time and there's almost nothing left to explore!"
5 - power regression based on eyeballing this curve.
6 - I can think of one better way, but it's going to have to wait until I get my copy of CoS back from a buddy.
Best Answer
Sorcery Points and Spell Points are 1:1 Exchangeable; but be careful with this change
Up through 5th level spells, the Flexible Casting table and the Spell Points tables list identical costs for converting [] Points to Spells:
\begin{array}{r|l|l} \text{Spell Level} & \text{Spell Points} & \text{Sorcery Points} \\ \hline \text{1st} & 2 & 2 \\ \text{2nd} & 3 & 3 \\ \text{3rd} & 5 & 5 \\ \text{4th} & 6 & 6 \\ \text{5th} & 7 & 7 \\ \text{6th} & 9 & - \\ \text{7th} & 10 & - \\ \text{8th} & 11 & - \\ \text{9th} & 13 & - \\ \end{array}
So it is clear that when they exist as a free-standing pool of points, Sorcery Points are more-or-less equivalent to Spell Points.
... Except when they aren't. Sorcerers don't use Sorcery Points solely to cast spells; they also use them to fuel Metamagics and Origin features. I'm going to use the Shadow Origin Sorcerer as an example, but many of these issues apply to other Sorcerers as well.
At level 3, Shadow Origin Sorcerers gain the ability to cast Darkness by directly spending Sorcery Points instead of a spell slot:
Under the normal rules, a Sorcerer has 2 2nd Level spell slots, 4 1st Level spell slots, and 3 Sorcery Points. This means, if they converted all their spell slots into Sorcery Points, they'd be able to cast Darkness using this feature 5 times per Long Rest (2 * 2 + 4 * 1 + 3 == 11; 11 / 2 == 5). Under your rules, they'd have 17 "Spell + Sorcery" points (14 Spell Points + 3 Sorcery Points) enabling them to cast the same spell 8 times per Long rest. So already, we're seeing that this feature greatly empowers certain kinds of Sorcery Point-powered features.
Not that this should be terribly shocking; the Spell Point rules as-is permit the same caster to cast up to 4 2nd level spells per long rest, up from the normal limit of 2, so some degree of empowerment of these features is expected. But what about features that aren't normally spells?
At level 6, they gain the ability to, at the cost of 3 Sorcery points, summon a Dire Wolf with special properties. Normally, a Shadow Sorcerer could do that 8 times in a single long rest (3*3 + 3*2 + 4*1 + 6 == 25, 25/3 == 8); but also note that they couldn't do that 8 times in a row, because summoning a wolf is a Bonus Action, which their Flexible Casting feature also requires, meaning they'd have to summon 2 wolves in 2 turns, skip a turn to burn off a spell slot, then summon one, then skip a turn, then summon one, and they'd have to skip multiple turns once they got to their lower level spell slots.
Conversely, with your rules, they'd have a pool of 38 Points (32 Spell Points + 6 Sorcery Points), meaning they could summon 12 wolves with this feature, and they would be able to summon those 12 wolves in 12 turns, no need to skip rounds.
I know you mentioned in your post that you're not concerned with Balance implications, so I'm not going to spend time judging whether this is a good or bad thing for the relative power level of Sorcerers, but I do think understanding the implications of this change on a Sorcerer's non-spell-slot abilities is important. This change will give all Sorcerers greatly expanded access to their non-spell features, which could have significant impacts on play. A Divine Soul Sorcerer, for example, would be able to far more liberally use their Empowered Healing feature, bolstering not only their own spells but also any effects their allies put out. It would also allow all Sorcerers to more freely use their Metamagic effects.