When I hire people, I am looking for the best overall improvement in the existing group. I do not hire for the benefit of the person being hired; I hire for the benefit of "the company" which really just means for the benefit of everyone already hired; hire whoever, once added to the group, will cause the most improvement in the group.

To give a specific example, I have seen a group of men in their twenties significantly improved by the addition of a woman in her thirties. Put simply, they grew up a bit and started acting more like professional adults; the output of the group was improved more by the addition of the woman in her thirties, I believe, than it would have been had I added an individually more talented male very similar to the existing members of the group.

Maybe what I'm saying here is that the hiring criteria included "what effect will this person have on the group" and, if one accepts that race/gender/religion has an effect on the group, then it seems sensible to consider those things when hiring.

I believe the problems are in _excluding_ people on grounds of race/religion/gender and not in simply allowing those factors to be part of the selection process.

I actually agree with you, and the scenario you described is interesting and makes for a good example.

Of course, that is different than what the linked article is demanding, which is having more women on panels just because, as if it is an intrinsic good, and without considering whether and how much it would affect the quality of the conference.