why not allow answers (and encourage answers) much earlier
We need the right kind of "site building" activity in Area 51— People focused on building sites, not using sites.
[see "Our Solution" below]
Area 51 is the place where users can go to propose new ideas for Stack Exchange sites. There are no actual "questions and answers" in Area 51. That's important. If you are suggesting that answers be offered, those answers better be damned good ones, or our core mission has failed. The Internet at large isn't going to understand the distinction of "oh, this isn't a REAL site, yet."
So, how do we do that? You talked about "getting rid of off-topic questions." You'll need voting and flagging and editing and closing and reputation to drive it all. That sounds like a Q&A site to me. What you are suggesting, essentially, is turning a proposal into a pale, trial-version of a site to see if the site might work out if we actually created it.
There's a big difference between defining a site and using a site. Asking questions and expecting answers (and providing answers) is using the site. I don't want to create "Google bait" with half-assed questions dumped into what may turn out to be an ill-defined idea. That goes against everything that makes these sites great. Quality and correctness matter! So we have to add in ALL those tools and behaviors which vet the value of the information now available on the Internet.
Good-bye Area 51; It would no longer exist, in spirit. We'd simply be creating Stack Exchange sites based on the mere suggestion that a topic might be a good idea. "Wouldn't it be cool to have a site about [X]?" *Poof* Done. That sounds like the failed model of Stack Exchange 1.0.
But let's say we can overcome the "trial site" problem. What if the proposal did not work out? Would we simply delete it?
There are two major problems with this approach:
Users are incensed when we throw away "content." Can you imagine if we let people spend months (or longer) contributing hundreds (maybe thousands) of questions and poured their knowledge into answering them… then, if the "proposal" doesn't work out, we just delete all that content?
Remember the first days of Area 51? We found that people were using the voting very differently from how we intended. So we discarded three days of voting data. That led to weeks of protests from users claiming to be demotivated, angered at losing all that effort, tired of us wasting their time, etc, etc; All that for discarding three days of voting… three days.The site develops momentum before it is even defined. People would simply start asking questions before the site had any boundaries. That run-away momentum would preclude being able to set those boundaries. Ill-conceived questions would attract more of the same. After all, what would preclude others asking similar questions (your "Google bait") when they see the same type of questions on the site? What are people voting on? The quality of the question? The appropriateness for the site? Whether it is a good on- or off-topic question? That subtle distinction would be lost on the vast majority of users.
##Our Solution
To build more interest around site-building activities, we are adding features to Area 51 which allow communities for form around groups of proposals. The general idea is to group related proposals into "departments," much like you might group academic disciplines at a university.
This will make it easier to add more communication facilities to help these groups of proposals work together and decide how to best allocate the creation of sites between them. There will simply be more to do with more people to talk about these activities within the subjects that interest them. With more people gathered around related proposals (and more to talk about) communities will form out of common interests… building great sites around subjects that interest them. More details about how that is going to work as the development efforts continue…
That's how you attract the right type of activity.