The core problem this entire argument runs into from a validity perspective is that you cannot prove that discrimination in favor off good test takers is causin

The core problem this entire argument runs into from a validity perspective is that you cannot prove that discrimination in favor off good test takers is causing EVERY positive outcome. Promotions, employee reviews, even legitimate measures of job performance could be an outcome of discrimination in favor of test takers. You would need to setup a parallel society where those who were good at tests in school were lambasted by their parents, held back from higher education, had performance reviews and interview panels done by bad test takers, etc. etc. It’s hard to actually disprove that this argument is justifying discrimination on the grounds of the effects of the same kind of discrimination. The more I’ve dwelled on this point the more I’ve felt the modern notion of discriminating on the grounds of identity group membership can’t be proven to be any more arbitrary than discriminating based on test performance. It's a subjective preference probably driven by direct or indirect self-interest.