That argument applies to the general population maybe, but that doesn't mean it also applies to the population of Google employees. Everyone there passed their hiring standards, so everyone has proved their individual merit already.
Your conclusion does not follow from the premise. It's not hard to come up with a distribution that is identical for the top x% but differs for the general population. Or one where the difference in top x% is in the opposite direction of the general population (e.g. two normal distributions where mu1 < mu2, and sigma1 > sigma2).
Well, again, the Google memo started by arguing Google's technical standard was being hurt so James doesn't seem to believe his colleagues all got the same interviews he did.
I'd be more inclicned to this sort of argument for racial and gender diveristy if Google & others didn't have such abysmly skewed distributions towards white men. Go look at Google's diveristy report and then do some Bayes Rule hacking with it.
Google's diversity rates are worse than the graduate rates for most schools I've checked. I have not found a composite data source to compare a larger average to.