One such experiment is WeKnowWhatYoureDoing.com - the Atlantic gives a good overview of how this works:
Let me give you this hypothetical about privacy. You and a friend walk into a public subway station having a conversation about how much you hate your boss. Someone happens to be recording every word spoken listening for the search string "hate my boss" while running facial recognition software to figure out who you are. This information is then being posted in another public location for anyone and everyone to see. Would that be OK?
My intuition is that almost everyone reading this post would say no. And yet, that is precisely what the website, WeKnowWhatYoureDoing.com is doing with public Facebook updates. The site scrapes public Facebook updates and searches for people saying "hate my boss," discussing doing drugs, giving out their phone numbers, or complaining about being hungover. It then handily formats them for broader consumption.
Interesting isn't? If you say something publicly, can you ever expect privacy? Is privacy really dead? Is Facebook then indeed really public (I say YES), and if so, why does it pretend to have privacy?(money, money, money - imho)
Then we have this experiment with twitter - a snapshot of twitter around the world in realtime: http://aworldoftweets.frogdesign.com/ Somehow this seems less threatening, perhaps, because, it does not focus on individuals... but it is using real time information from twitter users.
So, what makes one cool and one kind of scary? I would argue, the perception of privacy and the false sense of security that some feel with Facebook.