Boston Globe has an interesting article about Open Science, citing the routine list of worries that usually get associated with this idea, e.g., :
But in the world of science - where promotions, tenure, and fortune rest on publishing papers in prestigious journals, securing competitive grants, and patenting discoveries - it's a brazen, potentially self-destructive move. To many scientists, leaving unfinished work and ideas in the open seems as reckless as leaving your debit card and password at a busy ATM machine.
But, as John Hawks says:
I think that's a pretty simplistic rendering of how scientific credit is assigned. It ignores all the factors that depend not on your results but on networking. Who you know may be vastly more important than what you do.
I think that if more researchers were independent (not tied to someone else's lab) and if they spent less time grant-writing, we'd see more open collaborations. Right now the biggest barrier to openness is centralization.