I’m guessing that there is a strong correlation between subject matter Achievement tests and SAT scores. I’d be all in favor of just using the AT as a cost saving measure. Of course , then there’d be those who would try to prove that the ATs were culturally biased and force revisions to provide more equal outcomes rendering them less useful.
Murray makes the point that there is a strong correlation between parental intelligence (its heritable) and test performance. People who are smart generally (not always) do well economically. And it is difficult to make the case that math, physics etc. are culturally biased - though I am sure that our tax dollars are being used to attempt to prove that non-fact.