This kind of news makes me wonder if maybe we should just give up completely on patient privacy & by default just share everything.
http://michaelnielsen.org/blog/michael-a-nielsen/
At any rate, the folks on PatientsLikeMe.com are sharing, and they have results to show for it:
http://www.ted.com/talks/jamie_heywood_the_big_idea_my_broth...
Really inspiring stuff, because I assure you that the folks participating in these programs do not consider themselves hackers OR radicals.
Creative Commons has started an organization called Science Commons that has done some great work on getting scientists to share their data:
A promising step in this direction is Sage Bionetworks, which is attempting to use an open source style approach to drug development:
There are some very well known people behind Sage, and it's received backing from Merck (and Science Commons), among others. See, e.g., http://creativecommons.org/weblog/entry/19646, which states that Merck has backed them to the tune of 150 million dollars. At the moment they're just getting off the ground.
A much broader overview on open data and related issues can be found in Peter Suber's open access timeline: http://oad.simmons.edu/oadwiki/Timeline
Another interesting step is the Federal Research Public Access (http://www.arl.org/sparc/publications/articles/FRPAA-introdu... ), which would make most US-government sponsored research publicly available. It's not quite the same as open data, but is very consonant with it.