Well there's a subset of questions on TakeOnIt where the correct answer has a financial reward/impact. An example of such a question was Is there a housing bubble in the United States?. These type of questions overlap with the kind of questions seen on prediction markets (which is a nice model for monetizing intellectual content). I'd be curious as to the relative accuracy between prediction markets and using collaborative filtering on expert predictions.