That's another way to do it, so long as it's pretty codified and with low penalties, but I'd rather a system with more outright fiat on the martial end. You know - Wizards cash in fiat tokens to paralyze monsters and make fireballs appear, Fighters cash in fiat tokens to blind them, knock them down, etc.
There's a few reasons I think this works better than a codified "called shot" process.
One is potency. Blind is a heck of an effect, but one that's totally feasible for a weapon-user. If you're inflicting big attack penalties for trying to use it in a called shot scenario (and you'd almost have to to make sure it's not successful every single round), you're actively discouraging its use. If there's a token of sorts to cash in, you can avoid the penalty. More or less, if you make it too easy, you open up stun-locking. If you make it too hard, you might as well not have bothered.
Second is reliability - Buffy (to keep using this example) does crazy status effect stuff even to the big bads she fights. She puts them off-balance either physically or mentally, distracts them, knocks them back into obstacles, etc. In D&D, if you try to model this with attack penalties, you're losing out on this because that big bad is probably already pretty hard to hit. You're relegating "cool stuff" to attacks vs. mooks, and I don't think that's particularly satisfying.
One approach to "token fiat" is the 4e/Bo9S way or in that vein. Another is a more flexible Expertise Dice system. A third is kind of like Iron Heroes, where you build up tokens over combat and can cash them in for cool stuff. There's more possibilities, I'm sure - but I don't think a list of called shots is sufficient.
-O