A quick note on the tension between sharing everything as quickly as possible and keeping things for yourself.
The thrill of collaboration when like minds come together to brainstorm and solve big problems and the egoboo of having something you created "liked" or reused should not exclude or overshadow the value of figuring things out for yourself and having something you can point to as your own.
I think it's important to go offline for a bit and to work things out for yourself. Certainly, if you're reading something in math or science, you might try to work through the problem on your own prior to reading how the authors say to do it. I'm an extreme extrovert so I think by talking and writing (that's why you - like my husband - might be maddened by the apparent drift in my "convictions" or point of view). Others get the data, then go off somewhere and come back with an idea fully formed. What seems like ages ago now, I proposed that blogs were good to help people of these two groups work together, but I wonder about the pace of friendfeed and/or polymath projects and the necessity to feed the beast. How does that work for the introvert types?
Likewise with open science, perhaps, for theoretical scientists or for folks who need to go offline and then present ideas fully formed. Having someone jump in to their thoughts and tell them that they made a misstep in their proof or to tell them the answer instead of letting them figure it out for themselves, might throw them off their game.
Seems like for some projects the ideal limit of openness might not be real-time, complete, but at turning points when various milestones are passed... showing the work, but only after it's done and cleaned up. What do you think?