Not sure if you're joking but it would be a heck of a lot more fair. Just have some sort of objective bar like SAT score, and then do a lottery.
Then you don't have the "did we do enough for XYZ group", the RNG will simplify both decision making and justification. If there's x% of XYZ passing the minimum score, that's how many will be in your class.
"We didn't make the test. The bar is lower so more of every group passes. Fate decides who gets in, we don't add or remove anything based on legacy or donations etc."
In the cases cited in the article, I strongly suspect “zero decimal places of percentage” is sufficient to show discrepancies (that discrepancies of full percentage point or more will be present, given the admissions rate for one group changed from 73% to under 50%).
The article calls out the TJ admissions process as discriminatory.
The activists that are regularly called racist against asians proposed exactly this policy. It was called a "merit lottery" and worked by setting a GPA lower bound and then admitting students randomly if they had a GPA of at least that level. This policy was ultimately seen as "too extreme" and non implemented but it saw its fair share of "lefties hate asian success" and proponents were even called "enemies of excellence" by Scott Aaronson.
But if there were racial disparities in your primary and secondary education systems which lowered the scores of some racial groups, then your supposedly objective "minimum score cutoff" would still be discriminatory.
Then you don't have the "did we do enough for XYZ group", the RNG will simplify both decision making and justification. If there's x% of XYZ passing the minimum score, that's how many will be in your class.