[tptacek]

There are a huge number of assumptions, many of them quite dubious, encoded into the idea that this research is impactful to affirmative action.

Affirmative action might be problematic for totally unrelated reasons. Maybe there's no way to do it without being racist towards Asians, or without selecting a tiny privileged cohort of Black families to accrue benefits on (simply "applying for college" in the first place situates most people in a position of privilege).

We absolutely do not know enough about the psychometric science here --- forget race entirely and just try to pick through theories of intelligence and its mutability --- to make public policy decisions. It is, at present, a pure scientific concern.