He didn't just argue that there was a difference in the distribution but that the underlying cause was biological. Companies like google hire people who excell at certain skills and are therefore at the tail end of the corresponding distribution where the differences in likelihood occurence are largest. If he believes that for engineering there are 10 men for every woman at both tail ends of the distribution then the "natural" gender ratio in engineering teams must be 10:1 or standards are lowered in the name of diversity.

The implications of that worldview are totally toxic.