Size is not really the point here as far as i see it (the actual machine running this is actually quite a bit bigger than a classical computer - disclaimer: i have not seen their experiment, but other ones that go in that direction). The point is that if you want to do quantum computation, you need quantum objects to do your calculations with -- something an atom is, and a huge piece of silicon generally isn't.

Yes and no. Researchers are all too often playing into the hands of the current publishing status quo. There are quite some people aiming for high-profile publications (at least in my field, quantum physics, that is) that do not publish their best works in open access journals (or the arXiv), since the terms of the journals people like to publish in can be restrictive in that matter (or they just don't care for this possibility very much). That's the irony IMO -- people agree that for the sake of science and knowledge open access is much more beneficial than traditional publishing, whereas still scientists support the old scheme for the sake of having their name on publications on highly ranked journal papers. (Which, from a purely egoistic point of view, I can understand, by the way.)

I would go so far as to assume this might, on a statistically relevant scale, have to do not only with people's abilities, but also the teaching they got, which brings us to

That's quite an assumption there, isn't it? And comparing people's characteristics by nationality(!) (or even only country of residence) is hardly racism (unless the reader really refuses to see any other motivation, for there are many that are more plausible, having to do nothing with race).