This isn't a weakness in crowdfunding, at least not intrinsically.

Validating the team has always been important, but the general mainstream internet doesn't realize that yet. They will, after at least one highly public meltdown where millions disappeared down a drain for little to none of the promised deliverables.

(I hate to be negative, and I want to be wrong, but my money right now is on the Ouya - unproven team out of their depth promising an extremely complex product delivered at unprecedented low cost and unprecedented fast schedule)

I suspect soon we will see a redirection in crowdfunding and more focus on the ability for funders to assess the competency of a team, and I suspect we will see more credibility-building on Kickstarter projects than we have seen thus far.

That is a better way of phrasing my concern. I'm wondering if there is a way to build a system along the lines of a web reputation type thing which could identify people who could execute well. I could imagine something like:

  This team has:
         delivered 3 kicstarters (a, b, c)
         failed to deliver 1 kickstarter (d)

Of course that information will get accumulated and used but I'm thinking of a formalized way of doing that like ebay seller feedback or something.

I'm not sure I follow that. Is it a weakness of the model to fund "bad" projects from money that in the previous regime would never have gone to "good" projects? It's not from Kickstarter's point of view, obviously; they get paid. Ditto for the projects themselves.

The members of the public who are "investing" are obviously accepting risk. And I think it's reasonable to argue that the per-dollar risk is higher for these projects than traditionally funded ones. But that doesn't seem to be deterring anyone, so I'm not sure it constitutes a weakness of the model either.

Honestly, I think the only way to interpret your point is sort of uncharitable: the people in the most danger from the kickstarter model are the existing investor class, who risk seeing some of their likely prospects get money from elsewhere. I'm not sure that's really a bad thing.

Sorry, I wasn't clear. The challenge is that people who often invest (either as Angels or VCs or even large charity donations) learn through experience that the team that is going to execute the plan is the 'high order bit' or most important part of the equation. People who don't invest a lot, or are new at it, get caught up in the idea part of the pitch and imagine a world where that idea exists.

Inexperienced investors invest in the idea.

Experienced investors invest in the team.

The Kickstarter model opens up a source of funding for lots of new people, and it enables people who could not (or had not) previously invested in those people. That creates an environment which is exceptionally prone to failure. The education process will be a harsh. Because people rarely blame themselves for not thinking about the problem correctly they will start blaming Kickstarter, or the teams, and some of those people will do great harm to the system that is helping people do stuff they couldn't do before. That is why I think it is a weakness of the model, it doesn't surface the root causes of failure easily.