One of my concerns with this whole crowdfunding model is that if one limits the cap, then surely the expected value of the the returns from crowdfunding will be

One of my concerns with this whole crowdfunding model is that if one limits the cap, then surely the expected value of the the returns from crowdfunding will be negative. You need the gains from the uncapped hits to compensate for all the failures. If your downside is potentially losing 100% of your investment and that odds of failure are somewhere (arbitrarily picked) between 70-90%, and your upside is only a 20% gain 10-30% of the time, the math doesn't make it seem like the model is sustainable.