Even if we don't have total agreement on a formal criteria, I think it would be good if we could put together a document of recommendations in the form of "plusses" and "minuses."  TSC members might not agree on the importance of all of these, and each project being proposed for incubation would likely not have all of the "plusses," but hopefully we could come up with a criteria such that a proposal that had most of the "plusses" and few of the "minuses" would be overwhelmingly likely to be accepted, and the converse would also hold true.  This would let proposers make informed decisions about whether to submit a proposal or wait.

In the same vein, some guidelines suggesting whether a contribution would be more appropriate for labs or incubation would also be a good thing.

Some criteria proposed so far:

• I proposed that projects should have contributors/maintainers from two independent entities.  So far, we have never accepted a project that has not met this requirement (see Tracy's comment), and this ensures that companies submitting code have made steps towards open governance.

• Danno proposed that projects should be open-sourced in some form for a time before being accepted into incubation.  

• Angelo proposed that projects that propose new protocols should have appropriate backing in the form of academic publications.

I probably missed some things, so please feel free to add/comment (or fix what I've written above if you are Danno or Angelo).