I agree. Research and data might make people feel bad, and not making people feel bad is obviously way more important than increasing knowledge.
> We shouldn't accept these disparities as driven by natural forces until we've tried everything we can imagine to try to bring the differences in line.
I agree. Personally I think we should withhold all math education to males until they are 21. We can discontinue this policy when there is a 50/50 gender split among all STEM workers, educators, and investors.
> we should assume and act as though differences are due to bias long after the differences are well proven to be natural
Yep. 1+1=3. Doesn't matter if you can prove that 1+1=2.
> It is a case where being wrong in one direction is not very costly, but being wrong in the other direction (and thus institutionalizing bias) is disastrous.
Yep, unimaginable catastrophes.