His causality is off, way off. That's really the main issue. Both parents here have the same motivation, admirable enough, but one chooses to subscribe to an unreliable predictor for his decision making.

A predictor that is toes a really fine line between statistical reliability and outright bigotry.

In any case, we know that socioeconomic status is a far better predictor of educational success than race. There is a Hugh correlation between race and SES but it's far from causal. I think the main objection here is that he seems to claim it is.