The Skills Paradox: if your game includes the concept of 'skills', then what you will immediately do is guarantee that your Player Characters will not be 'skilled' in the way a real person would be. This will be directly proportional to how granular you make the skill system. The more granular, the worse it represents people.
On the other hand if you put no skills at all in, you will also eliminate skills in the much more practical way as GMs then have to do all the heavy lifting and players blindly argue what the characters are good at and not. So basically there's a continuum. and it has two axes, like all good barbarians.
The first is the broadness or narrowness of the skills themselves. Skills that are too broad get to the point where they look more like attributes or possibly even broader. "Knowing anything" is too broad of a skill category. On the other extreme end of the spectrum are skills so narrow they are possibly never used, the archetypal example is underwater basket weaving.
The second is the quantity of the skill system and scope of it. While this can be and often is married to the broadness and narrowness of the categories, you can have a narrow system with narrow scope, that meaning the few skills for example only cover things occurring in the dungeon (No Ride, Etiquette, Dancing, etc). Often D&D games have their skill systems stop short of combat skill (with some cheating here and there, like 2E's Blind Fighting and Tumbling NWPs).
Basically you never ever want to be a broad scope skill system with narrow focus (and it's pretty impossible to be a broad scope system with broad focus).
In the end it's a lot of work for basically nothing.
