As I mentioned over on Midas Oracle, I think this is a good approach because the ideas are assessed and developed before they are forecasted in the prediction market. That's the separation I think is important.

I think that when it comes to generating ideas, you need to be as open and inclusive as possible. The process should allow anyone that submits or helps develop an idea to share in any rewards from that idea. Once it's developed, then it can move to a stage where you can do forecasting via a prediction market.

Using a prediction market too early can do two things:
1- Poor forecasting due to social influence.
2- Limit revolutionary new ideas. It's too easy to short an idea that looks strange, when in fact it looks odd because it's revolutionary. The idea process should foster and develop ideas, not make them compete against each other.

I'm glad to have sparked a little discussion here.

I cases such as Rite-Solutions (and I realise that may be a particularly positive case, and that the company might have something vested in successful idea market stories seeing as they sell them) it seems that Idea Markets as an event can trigger communication about change within the organisation.