Okay, I know I've told you this like a dozen times on other occasions but I'll say it again: the real issue really is not optimizers. Believe me, the serious optimizer in my group hasn't said a word about 'feat taxes.' Like you, he just accepts that some feats are stupidly good compared to similar feats, and he takes them.
The real issue is that some of us simply want the game to be played by the design tenets that we like, the design tenets that the devs promised us: that no one option should be stupidly good compared to other similar options. That nobody should have to take any specific feats for the game to play as intended. The devs delivered on their promise with regards to 95% of the game, which is why 4e is my favored edition. It's just that last 5% that irks amateur devs like me [and at least one professional dev].
I don't see any good reason to not fix the math holes with true errata, especially in an edition where errata is such an accepted part of the game. Even if the "we fixed it wih feats so that all players can take advantage" line is true, it can be fixed with errata at the same time. (Hint: feat bonuses don't stack!) I've accepted that it's never going to be truly officially fixed, or even widely officially recognized--until 5e gets announced, possibly, so they can use the math hole problem as a selling point for 5e--but I'll always think of Expertise & Family as feat taxes.