Reading Soren Johnson's blog on games teaching lessons there's another point I'd like to bring up.
I think almost all games, including WoW, teach one really important lesson, which I'm having trouble putting into words. So I guess it's good there's games. But I'll try anyway. Any game has an optimal strategy. Finding that optimal strategy requires time, research, practice. It often isn't what you expect or want it to be, no matter how clever that imagined strategy is, or how deeply you think you understand the game designer. (I *so* wanted hammer-ons & pull-offs to be a valuable skill in Guitar Hero, but it seems I have to face the truth that they're very rarely useful. Course, I still haven't beaten that Freezepop song on Expert.) The strategy usually doesn't translate from one game to the next. Although most of the 'lessons' you might learn from simulation games (time is more important than skill [from WoW], you shouldn't have a child unless one parent works [from The Sims], or line-extending a brand is a bad idea [from Capitalism 2]) don't necessarily apply in real life, this sort of meta-lesson does: you get better at something with study & practice, your initial guess at a best strategy is usually wrong or at least needs major tweaking, and strategies that work in one context may not work in another.