I'm not familiar with the adventure path you're using, so I can't speak to it specifically, but I can give some general advice.
There are, generally, four ways to handle this.
As you mentioned, is let them walk in and die. They'll,
hopefully, learn from their mistakes and be more prepared in the
future. A lot of people like this kind of play and it's perfectly
valid, so long as they have a way to gauge the strength of the area
they're going into or some warning that they might be in over their
heads.
However, some players don't like having their characters die
and some groups would rather not have to deal with the
discontinuity. It can also be a bit immersion breaking for games
that are more focused on story, rather than just tactical and
mechanical prows. You said that you have younger players and they
don't like their characters dying, so it might be best to avoid this method.
Come out and warn them outside of the game. "Just a heads up, the area you're going into is meant for higher level characters, so you might want to come back later." You should still give them the choice to go in, if they want to, but then the decision will be entirely on their heads and they'll also be able to react appropriately.
This method has the lowest chance of ending in tears and the best chance of ending in success, simply because it's straight forward and gives the players the information and opportunity they need to make an informed decision. That said, it's also the most meta-gamey and that can be a turn off for some.
You can also edit the campaign so they simply can't get in before they're ready. You said there's a banshee in the canon version that can tell them what they need? Ok, well, in your version she doesn't know. Simple as that. If they've never played the adventure before (and haven't read ahead) then they'll never know they missed anything. They won't learn anything, but they're guaranteed not to get to the cave too early.
Don't worry about it. If they get to the cave early, make sure there are some hints that they might be getting in over their heads, but let them go. Then if and when they get into a fight they can't win, let them run. You can also arrange the fight to minimize the chance of them dying, without going easy on them, by doing things like having the monsters prioritize still standing characters over fallen (but still living characters) and active combatants over characters that are trying to evacuate the fallen. If there's something about the cave that means they can't exit it once they enter, you could put a comparable fight (maybe a little bit stronger than average) at it's entrance to serve the same purpose.
This method is the most game friendly, but it also has the highest chance of someone dying (baring the one where you just let them all die).
Personally, I prefer the fourth method, since it's the most RP friendly and knowing when and how to retreat in an RPG is a good skill. I don't think the players should always win, but by the same token defeat shouldn't default to them dying. There's also the possibility that good tactics will make up for the level difference, in which case they'll feel really good about overcoming a more difficult challenge. Even if they retreat, getting stronger and coming back to beat someone who's already beaten makes for a great sense of accomplishment.
That said, if you're really worried about them dying and you think they won't handle it well, #2 and #3 will serve you quite well.
Metagaming
First, lets kill the metagaming ad hominem: "Metagaming is any strategy, action or method used in a game which transcends a prescribed ruleset, uses external factors to affect the game, or goes beyond the supposed limits or environment set by the game. Another definition refers to the game universe outside of the game itself."
Optimising a combat-oriented character to be good at combat within the rules is not and never can be metagaming. You can't even mount a game universe argument that it is: a person who has devoted his life to being a wilderness warrior (aka a Ranger) is going to learn to be good at fighting or die!
Comparison
Hit Points
Your hit point edge is insignificant; an 11 hp advantage is, on average, 2 hits or 1-2 rounds more staying power in a combat (less if fighting multiple foes). When you consider that the Paladin has an ability to heal 15 hp with their Lay on Hands ability at the cost of an action, they effectively have more hp than you do. You do have a definite advantage if you are being hit by things like fireballs; on failed saves you are the only one left standing.
This is an edge but a small one.
Damage output
I will assume everyone has the same stat modifier on damage rolls.
If you are using your bow and choose to use a spell slot for Hunters Mark, you can do 3 + d8 (bow) + d6 (Hunters Mark) (avg 11) on the first hit and the same plus d8 (Colossus Slayer) (avg 15.5) on subsequent attacks. This is great if you are fighting a monster with lots of hit points; it is not so good against a dozen goblins since the first hit will drop them and your Colossus Slayer never kicks in.
Meanwhile the Paladin with a longsword and the dueling fighting style is doing 3 + 2 + d8 (longsword) + 2d8 (Divine Smite) (avg 18.5) (I haven't considered some of the really cool spells they have).
The Rogue is doing 3 + d8 (longbow) + 2d6 (sneak attack - a good rogue should almost always get this) (avg 14.5).
The Sorcerer has a plethora of options (Magic Missile, Burning Hands, and Cloud of Daggers spring to mind) or they can just fall back on a damaging cantrip for d10 (avg 5.5). If they are a gambler, Hold Person can end a combat with a single humanoid on one failed saving throw.
If the Bard wants to be handing out massive damage in combat then they chose the wrong class; that is not where their talents lie, they are an enabler - they enable others to do more damage.
The Ranger is not the best at handing out damage.
Overall, you are playing your character to his strengths; are the other players playing to theirs?
Pacing and Encounter structure
You say "I was typically able to go first in any combat due to high DEX, and dealt such insane damage that the guys going last did nothing".
I read "The encounters are underpowered".
Don't misunderstand me: it is the nature of RPG that the PCs will win (almost) every fight because they can only lose once. Most combats will be and should be cakewalks, they are there because combat is fun and they consume resources. That said, they shouldn't be so insignificant that they are over before the first round ends. A quick combat like this is great if the players have planned and executed a great ambush, its not great if it is just way underpowered.
If you have enough spells to use a spell in every combat then you are not having enough encounters between long rests. Burning through spell slots for a non-core spellcaster should be a tough decision: "Do I use it now or will I need it latter?" If you are not thinking this, at least briefly, all the time then your DM is being easy on you. Fights early in the day will usually be easy but this is due to everyone having lots of resources, as you burn through spell slots and hp the same encounter becomes much harder.
Also, the structure of encounters matters. 5 PCs on one monster is an easy fight (unless the monster's CR is extremely high for the party); the monster can only target 1 PC while copping damage from all 5. 5 PCs on 5 monsters is much harder; the tough PCs have to control the battlefield or the squishy PCs will get squished. 5 PCs on 15 monsters, even very weak monsters, is really hard; everyone is copping damage and the fight will last 4-5 rounds minimum.
Best Answer
The only limitations mentioned in the spell description are:
Other than these explicit limits what the spell works on is decided by you as the DM. If you think some damage is too extensive for the spell to function then it won't work. You can also talk to your players about setting reasonable limits for issues the rules are not clear on.
As for your examples I would rule the spell doesn't in the rust monster case and can work in the cloak case. The spell mentions breaks or tears and I wouldn't consider rust damage either of those things. For the cloak I would rule that multiple castings of the spell can be used to repair multiple tears.