Corporate prediction markets

There's been a lot of noise recently about companies using trading markets to help predict outcomes, ferret out the best ideas, or accurately size up a complex situation. BusinessWeek Online packaged a whole issue on it, and Business 2.0's September issue (p. 47) discusses "The Wisdom of the Corporate Crowd." Yes, this draws on that wonderful pop psychology of The Wisdom of Crowds. This is a business process that I see as one part interesting solution, one part ugly band-aid. Yes, it's very interesting to see how, as Business 2.0 points out, a corporate market can 40% more accurately predict HP's montly revenue than the official forecasts. However, most of the solutions seem to be applied when the betting pool is a workaround for an organization's ugly tendancy towards groupthink. Just as the Busienss 2.0 article calls out, "Usually it's the loudest, most obnoxious guy who gets heard." The other ugly aspect of such markets is how easily they could be misapplied. My guesses of the most obvious mistakes in integrating them:

• Unwise participants. The participants in the markets really don't have good or better insights into what they're betting on. Take for example the "terror market" proposed by the Pentagon; few if anyone in the public has more insight into terrorist activities than does the Pentagon, so it's more guesswork than averaged out wisdom. Compare that to the stock market, where detailed knowledge of markets and firms can be applied by many of the participants.
• Little or no incentives. If the participants aren't going to get some payback for their insight, they won't be working very hard to investigate and carefully consider what they're betting on. Without incentives, you're dumbing down your crowd.