I would guess that there are lots of places where having one gene be allele A and another allele X would be good or B and Y but having A and Y or B and X would

I would guess that there are lots of places where having one gene be allele A and another allele X would be good or B and Y but having A and Y or B and X would be bad. And I'm not sure if current screening methods would find that except by saying that A is good if X is more prevalent in the population than Y.