Every game has that the player behind it makes a huge change in effectiveness. D&D combat is no different that D&D roleplay in this aspect. A player making good tactical decisions, analyzing the foe and the environment, and acting as a team player will be enormously more effective than one who just seeks to do damage to the closest opponent. And that's before the player input to the character build, where combat-related is like 80% of the mechanics of the character sheet.
Now, that doesn't mean I don't have suggestions. And they are around that characters can do things the players can't. It makes no sense that a character that can, for instance, fly even though the player cannot, can't give an inspiring speech where the player cannot. So a character needs to be effective at everything their character sheet says they are effective in, even if the player is not. A player who knows nothing about woodland survival, and in fact makes poor decisions like suggesting a camping spot where it will likely get flooded on a rainy night, needs to have their result elevated to the character's skills, which would avoid this. You wouldn't penalize the wizard because the player narrated some really bad faux Latin when casting, this is more of the same. The character is the foundation of competence.
But playing well should have a positive effect. Be it strong tactics, an intelligent search plan, or framing something to be persuasion vs. deception when the character is better at one than the other. And there's an interesting place for your traits - should a silver tongued player be able to deceive someone when the character has no skill, or perhaps a negative. It may be that having a trait like Uncouth would not just have a mechanical effect on the role when it would come into play, but also limit the best effect so that a player can't exceed it through their own skill.
I don't know if there's an answer for that which will satisfy all players.