No, they are not, but is it a good thing if higher percentages of lower scoring applicants flunk out? And if they're graduating at the same percentages as higher scoring applicants, I'd like to know why, because then high scores don't really predict academic success and we need to figure out what does. We don't do any student any favors when we shoehorn them into programs they can't successfully finish.
My generation didn't have pre-school, heck, I didn't have kindergarten, yet they had to recalibrate the SAT for the next generations because their scores kept dropping. More school isn't always the best answer, better school is.
Not ignoring them, steering them into schools that are more likely to see them graduate into successful careers. I wouldn't steer a rich white kid into the top elite Ivy League schools unless they showed the aptitude necessary to excel, because every kid who gets into a program he/she doesn't deserve knocks another kid out who could probably do it better.
Can you quantify what you mean by "the outcomes are better when we implement AA"?
If they pass the same classes with the same requirements, absolutely they should be assumed to be competent doctors. If, however, they "pass" with lower scores than those required from students with different colored skin, they haven't met the requirements. You might as well lower the physical requirements to be a fireman so wheelchair bound applicants can be accepted. Sure, they have the coat and the hat, but you wouldn't expect them to carry an unconscious person down 5 flights of stairs.
And, I have to question your assertion that more black doctors means more doctors in Detroit. Black doctors are just as likely as white doctors to choose more lucrative, less dangerous places to live and work.