My motivation for asking this question was to get people thinking about whether sites outside the main computer/technical/science domains of existing StackExchange sites needed additional criteria to ensure they had the necessary expertise committed before going into beta. Sure, it is a worry if a site about human sexual relations has very few women participating, unless of course it is a site targeted at gay men.
But it is more worrying if a site attracts only 2.5% Experts out of 200-odd committers.
The existing expertise designations that people choose when they commit might not be a perfect indication because they would be gamed if they became a criterion. But I think we (or the moderators) need to have a conversation about how we gauge whether sufficient expertise has been assembled to create a viable site, for domains where C# expertise or whatever got you the rep on the StackExchange is not a good guide to your expertise about this topic.
This is in some sense the converse of the issue the Libraries and Info Systems proposal has - lots of expertise but not enough StackExchange rep.