My broader issue is that Kickstarter, in general, seems to have a dilemma on its hands. On the "buy" side, it has a lot of users who are essentially treating Kickstarter campaigns like product preorders. On the "sell" side, it has a highly variable mix of project owners. Some are doing presales. Some are seeking funding for proof-of-concept. Some are seeking seed money. Some (seem to) have finished products, and are evidently using Kickstarter as a marketing tool. Etc.
The challenge is that you have a fairly homogenous use case on the buy side, and a heterogeneous use case on the input side. Adding to this challenge is the fact that the use case most "buyers" seem to be gravitating toward is directly at odds with the use case Kickstarter wants to encourage on the input/"sell" side.
Now, my recommendation stems from my assessment of the problem. I may be wrong. But to me, when I see a lot of variance on the sell side of a market, and considerably less variance on the buy side, I tend to think the solution involves bucketing, or categorizing, the different product types on the sell side. The goal is to create some sort of homogenous subsets of what is currently a very messy, variable master set.