But we haven't settled whether they're being racially discriminated against. If some legitimate feature to discriminate or filter against correlates with race, then it will look a lot like racial discrimination when it isn't. Harvard wanting to maximize their future prestige by carefully constructing the makeup of their student body is such an instance. Maximizing leadership potential of their student body will look like racial discrimination. It's similar to how the distribution of housing loans can look like racial discrimination if you didn't already know that income correlates with race.
> non-Asian leaders who are individually worse leaders take the spot of Asian leaders who are individually better leaders
No. I'm saying GPA and SAT scores only correlate with leadership potential up to a point. Once you've passed that threshold, other factors start contributing more, e.g. your background, interests, etc. It is not the case that the person with the higher GPA/SAT necessarily has more leadership potential.
And somehow these "lacking in leadership potential" students happen to be Asian more often than the norm? Year after year? Seems like a convenient coincidence.
Then, pray tell, which lack of "being an outlier in some socially/politically relevant cluster" is common to Asian applicants, year after year? I'm sure we would all be fascinated to hear the specific areas in which Asian applicants are failing to match these heuristics where other minority applicants do.
But to put a very fine point on it, Asians with stellar scores, similar backgrounds, similar interests, similar extra curriculars lack diversity in backgrounds, interests and extra curriculars by definition. There's nothing deeper to be noted about this. Whether Asians that apply to Harvard actually have similar backgrounds, interests and extra curriculars is certainly up for debate. But it is plausible.
The issue is that the distribution of leadership potential in the applicant pool is not equivalent to the distribution of top GPA and test scores (past a certain threshold). So when you use your GPA/SAT metric to analyze something that doesn't follow the same distribution, you will conclude there is bias. But it is your metric that is wrong. If your argument is that they necessarily should use your GPA/SAT metric, then you have to argue for this directly. It is not self evident.
What you're saying here is that non-Asian leaders who are individually worse leaders take the spot of Asian leaders who are individually better leaders because the race of non-Asian leaders makes them better leaders.