That rankles against my (and, evidently, a number of others') design sensibilities, though; at any one point in advancement, any two options should be mechanically similar in weight. With +1 to ability scores, half the time there's a direct mechanical effect, and the other there's a delayed mechanical effect. You see how that doesn't jive? I mean, if feats were balanced against the average of +1 to an even score and +1 to an odd score, it'd work out, I /guess/, but it still seems harder to design around than it needs to be.
It's the same problem with balancing a powerful end-game wizard by making low-level wizards pathetically weak - it makes assumptions about play that probably won't be borne out in actuality. Admittedly, it's less drastic than that, but it's still there.
You don't design the endgame, then assume the road to that point is free of bumps, y'know?