When an Area 51 proposal makes it to private beta, the site has 2-3 weeks to show viability or be closed instead of advancing to public beta. I agree with that, but I wonder if we're being clear enough about the expectations to the users of those sites.
We seem to promise that the site will advance to public beta:
The email that committers receive when the private beta starts says this (or did about nine months ago, the latest sample I have):
Q: How long is the beta?
A: After a few months, when it becomes abundantly clear that it's Making The Internet A Better Place, we'll slap a proper logo and design on it, and make it a full-fledged citizen of the Stack Exchange network.
The community members have just been through a months-long process, maybe up to a year, to get to this point. They probably aren't expecting to have only a couple weeks to validate the site.
Proposals recruit, and are sometimes proposed by, participants who are not all that familiar with the inner workings of Stack Exchange. We should not expect them to know stuff that isn't in the Area 51 FAQ or the proposal pages or the email we send them.
This question is not about lengthening private beta. It's about better communication about expectations and timeline.
There are no formal, objective requirements; the CMs make a judgement call based on their vast experience with network sites. They have some rules of thumb, and even if we don't want to be too specific, we could at least use some hedging language:
In this answer to another question, which is about public beta, I suggested some wording changes for that email that would also help with this issue for private beta.
On that Area 51 status message, maybe we should say the site will "be evaluated soon" or that "if all goes well, the site will advance..." or something like that.
We don't want to be discouraging and cause users to worry, but we do want to send the signal, somehow, that reaching private beta doesn't automatically mean a site will go to public beta. We hope it will, and the email gives some good advice about what to do during the private beta. But it sounds like we're making guarantees that we do not, in fact, make.
I don't know how often private betas are closed, but it does happen and each time there's a group of disappointed users, so it seems worth a small investment to be a little clearer.