Ross Mayfield summarises current discussions about the 'trustworthiness' of Wikipedia.
Which brings me to an lingering thought -- that explicitly codifying reputation introduces a cost which can constrain commons-based peer production. Wikipedia was never supposed to work, somehow does because of good club theory and transaction costs, and has gained a reputation as a resource. Introducing reputation for contributors or articles is the greatest risk to the Wikipedia community. Getting a base study on factual accuracy can help inform this decision as well as educate the public on how to use and participate with this commons resource. [Many-to-Many: Wikipedia Reputation and the Wemedia Project]