Question is off topic for this mailing list. Please direct to the dev list. P
I would try to identify goals that can be verified automatically,
rather than goals that require a lot of human involvement, review (and
I am mostly opposed to an approval process).
Jay's metric of having X people involved who have previously run a
P2PU course is a great example. It's clearly defined, and there is no
ambiguity. If we had more flexible metrics they would require a lot
more attention and time to review. (by the way - I think the magic
number X is 2, but that's gut feeling)
I also like Lila's idea of adding some measure of activity / demand.
In the same way that stackexchange requires new communities to show
sufficient energy before they get a free, hosted, Q&A site - we could
require schools to have a certain number of users who follow or
participate in courses, and a certain number of updates to
courses/groups affiliated to the school. Again, we can choose
indicators that are easy to track automatically to reduce the need for
manual review.
P
Number of participants is definitely a stronger indicator than followers, but there might be a place for both. I know there are probably lots of people (such as myself until this past cycle) who lurked around P2PU but didn't actually join in on any courses. So I think there's something to be said for getting a quick headcount of the people who say "yeah, this is a good idea" in addition to the people who, by participating in a class, have already expressed that they think it's a good idea. On the other hand, I could see why it might be too complicated to rely on two different but closely related metrics... But something like "has 100 followers and 25 past class participants" (or again, some other numbers, those are just placeholders) could work.
There is also the possibility that, depending on how we organize these stages, we might want these two different metrics at different stages. i.e. early on in the process we only check for followers, but in a later phase we check for active participants). This would actually be parallel to the StackExchange model that Phillip mentioned.
For those who aren't familiar, this link gives an ok overview of how proposed new Q&A sites work at SE: http://area51.stackexchange.com/faq - you'll probably only need to read down through "how do I start a new site?" as the rest is more SE-specific. (As you're reading, just replace "site" with "school" to translate it to P2PU.) Obviously it doesn't mesh with P2PU 100%, but it might be worth pilfering a few of their ideas, since they have been very successful in what they do.
Lastly, I definitely think microgrants should require a much stronger indicator that the school (and its organizers) are committed and able to follow through than the standard requirements for becoming a school. Grants represent a significant commitment of resources on the part of P2PU so I doubt every school will de facto qualify. I think, though, that microgrants are a place where it does make much more sense for human beings to put in legwork rather than relying on cold metric; much of determining who gets a grant will be based on how strong the proposal is and the trust that P2PU decision-makers can put in that school's organizers.