Unfortunately, with the case for perfectly effective altruism, the ambiguity of these type of arguments of where the "best" impact can be made usually fold back into the the idea that the donation is seeking not just advancement in the world (though this is clearly and hopefully part of it), but also to build a personal legacy and connection to things that the donor cares about - in this Stanford. We like to reinforce things we can see and things we care about. Empathy and investment exist in much narrower scopes that we would like to think. I agree with you that the $$ could be allocated perhaps more efficiently, but hey - at least he is putting it to something positive. If more and more billionares take a note out of Phil Knight's book (referring just to the donation) we will slowly make good (perhaps not ideal) progress.