I knew there was something I meant to include in the rant about game design, and then Timrod reminded me of something I've been thinking about since reading Spawn of Endra's post. A lot of the fiddly bits we come up with are intended to add realistic details to the game. (Another large chunk of fiddly bits are intended to "balance" the game, but that's worth another post...) And a lot of these fiddly bits suck because the realism isn't fun; it makes the game more tedious instead of more entertaining.
I drew the distinction in a comment on Spawn of Endra's post between quantification and functionality. If a rule adds more precise measurements of some kind, distinguishing materials by real-world density, flexibility, malleability, and brittleness, with modifiers based on manufacturing processes, technology available, and skill levels of craftsmen, in order to create really detailed weapon breakage rules that reflect real-world probabilities, that's just not fun. It doesn't mean that weapon breakage itself isn't fun; it's just that the way it's handled is tedious.
On the other hand, you can say "weapons ought to have a chance of breaking, although not very often. Just often enough that it's noticeable." And then you create simple rules that add that function to the game, without worrying about precision. And then maybe you tweak the rules to make them easier to use or easier to remember. There's nothing wrong with that. You can add to the game. You can go back and review what you've added, to make it simpler. Hell, I've already written a simpler version of the rules I wrote on deferred varieties. They'll be up tomorrow. Because I think they add a lot of detail without being too fiddly.
So there's no sin in adding rules to the game, as long as they don't double as your master's thesis.