Anybody know if someone is already working on making the datasets mentioned on the article easy to work with? Quick search for some, the look very interesting but a pain to get the actual data.