Two points:
1. For all the logical/social contortions, judgment calls, distorted selection standards, tuning, and subjective assessments involved to create the "desired" class of students (thousands and thousands of hours of applications, reading, discussion by admissions officers, students writing essays that emphasize their overcoming disadvantage)... how much better an outcome or result does this create than simply allowing some objective standard to dictate who gets in, or random selection above a certain bar?
Is there any way to say how much better we do for so much effort? Other countries apply strict bars for admissions -- do they do worse? Why do we think that this produces any much better outcome? Is it just to make us feel better?
2. By what principle are administrators of universities limiting their actions or extent of tuning? Do they apply this believed solution to university admissions just because that is their domain and have it under their control? What makes them think that this is the solution to inequality? And if this is the solution, why are they so meek about it and don't instead completely overemphasize every group that needs more representation?
How do they know when they've achieved the goal? How do they know whether a new goal needs to be set? When "equality" on some dimension is achieved, will that be the end of it? Is there an end to this? Will other groups be not so lucky to get such attention when their turn comes?
Unsatisfactory answers to all these questions make me not support how AA is implemented in the US.