Interesting parallel: Y Combinator is a breadth-first search through the possible solution space. It galvanizes a number of potential solutions (similar to following a number of promising concept designs) and eventually is successful.

So, in my experience, concept designs are best used when two wildly varying ideas are able to combine. Do two failed startups take their ideas, mesh them together, and produce one new totally innovative one?

I'd like to hear stories about that.