From the print edition of this New York Times article, I learned that at TradeSports.com, the futures contract for "Weapons of mass destruction will be found in Iraq on or before September 31 [sic]" went for 80 in March, meaning bettors thought there was an 80% chance of finding the weapons by September 30, but now this contract goes for less than 30, which means a 30% chance or less. (Here's the TradeSports chart.) So either the bettors were wrong early on or (if some weapons do actually turn up) they're wrong now. Either way, they've clearly been following conventional wisdom. Where's the predictive power of the market?
I will add, though, that Professor Robin Hanson, an advocate of such markets, does bring up a good idea in the Times article: maybe the government should construct a betting pool and limit it to intelligence professionals. This way people who have some expert knowledge can anonymously put forth their opinions -- which might not be the opinions of their superiors. (I do believe that line workers often know a hell of a lot that their bosses don't know -- and should.)