It would sorta make sense if in on-on-one, unarmed combat between creatures of the same size, that a wrestler couldn’t level up higher than, say, one level higher than the wrestling skill of the opponent, maybe. But it’s not as clear which skill you’d use for an opponent in more complex situations. Would you compare against the opponents armor user skill, or whatever weapon skill they have equipped?
Well, you say, we have to take size into account. But if you pick anything but the correct combination of factors, you just push the exploit somewhere else. Then we’d just have to wrestle large harmless enemies instead.
In this game, fighting five of the same enemy is way harder than five separate battles against solo enemies. Would you take that into account? And what would you do if some of the group were dual-wielding weapons, two of which are covered in slow acting poison, oh and also one enemy has a breath weapon?
So what we really need is some estimate of the total difficulty of a battle, basically a concept like an “experience point”. But that would ideally have to take into consideration such myriad factors as the terrain, that it might well be an intractable problem.
That’s the beauty and the curse of the skill-upgrade-on-use paradigm. It avoids the above conundrum, but it has exploits. The good thing about exploits is that you can set a challenge for yourself not to use them, so you can think of it like an Easter egg if you want to be really generous.