IMO it seems that if there was a good data driven way to measure performance, it would allow adjustments to optimize for positive outcomes. As it stands, it feels like the majority of diversity initiatives are evaluated subjectively instead, so diversity officers effectively act as a political role more than serving the intended purpose.
This NY Post article definitely has a bias they are promoting, and the Heritage Foundation (conservative 'think tank') which ran the study certainly does as well. However, I hope more similar studies are performed - quantifying the effects of diversity initiatives is crucial to evaluating which are productive and which are counterproductive. There is likely a bell curve of outcomes, and we don't know which initiatives fall where.