Between one person who is better at calculus and another that's better at chess, which is more meritorious?
Depends on whether he's applying to a chess or physics program, no?
After I tell you that merit isn't a scalar and invite you to consider measures of merit that are somewhat orthogonal to each other, the best you can come back with is to say it depends which (orthogonal) field of endeavor they're seeking to enter? Pretend the answer is neither, they're competing for the last available place in a medical school.
My whole point here is that facility with calculus or chess both demonstrate cognitive ability, but there's no universal standard for asserting that one is better than the other - we can only talk in terms of what sort of utility it is that we wish to maximize. Likewise, absent universally agreeable standards of merit, we can't easily say one group of people is better than another while neglecting structural biases. It might be that people who are good at calculus would in fact make superior chess players if we were to invest in some remedial chess lessons to rectify the shortcomings of their strategic education, so to speak.
I fail to see how this supports the claim that an admission system is non-meritocratic simply because some people are unable to pass it due to bad life circumstances.
The fact that our predictor is imperfect doesn't mean that it magically becomes better when we attach racial weights to it.
Between one person who is better at calculus and another that's better at chess, which is more meritorious?
Depends on whether he's applying to a chess or physics program, no?