I suppose, but they are only opening articles that are 70 years old. Which means curiosities like Darwin's geology papers are now available, but you still need to pay for anything remotely current. So it's a nice start, but unlikely to effect their bottom line, or make doing science any easier for anyone outside the first world academy.

Yes, you're right. But it's still a step in the right direction and hopefully others will follow suit. And I am really optimistic, looking at the uber-cool initiatives of the Khan Academy (http://www.khanacademy.org/), MIT's OpenCourseWare (http://ocw.mit.edu/index.htm), Stanford's OpenClassroom (http://openclassroom.stanford.edu/MainFolder/HomePage.php) and their latest online courses in Machine learning (http://ml-class.org), Artificial Intelligence (http://ai-class.com), and Databases (http://db-class.org).

They didn't have much choice in the matter. For a few months now there is a torrent floating around with most of this stuff in it.

While I think it's a bit extreme to say they didn't have much choice, it seems to me making data more open is becoming somewhat trendy and this sort of peer pressure may have certainly played a role in the Royal Society's decision. Anyways, the inertia of the status quo shouldn't be underestimated and although things are really not that fair, it will probably take quite a bit of time before we see widespread democratization of data.