> If that holds for Caltech applicants, which is reasonable to assume

Caltech is an outlier if there ever was one, so why would it be reasonable to assume Caltech’s applicant statistics reflect broad population SAT statistics?