I don't agree with others that it's only a matter of time - I think there are several hurdles here, that are needlessly inhibiting otherwise successful proposals from getting off the ground.
Summary
This has turned into a pretty long post - I'll attempt to summarise it properly later, but for now I'll go with this...
The hurdles for new Area51 users are basically: not enough relevant information, too much unnecessary stuff.
Scroll down for some pictures of possible solutions.
Hurdle 1
The first problem is explaining the benefits of StackExchange to someone that hasn't encountered it before - and indeed probably settled on the first place that Yahoo/Altavista/Lycos provided them, and simply put up with all the quirks because they don't know they can have better.
How do you explain the platform when you can't offer them any meaningful comparisons?
There are no active artist-based sites, and only Photo.SE in beta (still!?) which comes close.
Not all topics will necessarily have a related idea, and even then I don't think relying on this is the best way to convince someone that a StackExchange site is better - it needs more than that (will make a suggestion later).
Hurdle 2
The next hurdle is Area 51 itself - you've convinced someone and they've followed your link, but they arrive at a screen full of confusion.
It might not be obvious to practised Area 51 users, or even just people experienced with browsing lots of websites and webapps, but I imagine the first thought of someone following a link is "WTF is this!?", and "why is this different to <example site>?"
There's not really a clear explanation of what a proposal is, in a general sense, which there really should be for an off-site referral link. And the text in the orange box isn't helpful enough: "If you like the idea, you can click the "Commit" button below to be notified when the site is created" - there's no indication of what Committing is, or that there that the site will only be created if 200+ people commit.
Some people are scared of commitment, and may simply want to wait for the percentage counter to get 100%, but I don't think it's necessarily obvious that the counter depends on commitments.
Hurdle 3
The final hurdle is the Commitment box itself.
Again it's a confusing muddle: What's that red beep thing - looks like feedback - did I do something wrong? Why is there a huge highlighted 3?
Then it's asking people to almost sign a contract - people don't like being locked into things, but you're asking for a guarantee of participating actively for three months, and requiring a minimum of ten questions.
It might be intended to weed out less serious people, but it's as likely (or more so) to scare away people worried about being punished if they don't comply.
Solution 1
The first problem could be helped by generating a sample/tutorial site once a proposal reaches commitment stage, not functional, but populated with the proposed questions and random (lorum ipsum?) answers, and nice verbose tooltips explaining every aspect of the interface, and how it differs from traditional Q&A/forum systems.
This should be accompanied with a functional Meta site, where the proposer and any SE user with >X rep can discuss how to find committed people, and start on some of the early private beta meta discussions.
Solution 2
The second problem ideally needs a simplified referral screen - hide the users by default, give an explanation of what the example questions are (i.e. on-topic=good questions, off-topic=not what the site is about), and most importantly give a clear explanation: what is area 51, what is a proposal, why commit to it? (And of course a link to the example site described above, with a clear indication that it won't get created for real without commitments.)
Here's a quick mock-up of what that might look like:

( large version )
Solution 3
The final problem has an obvious solution too. Tidy the box up a bit, (simply by not putting non-significant red text in the top left, and removing the formatting on the proposal count).
Then, reword/simplify the opening paragraph, and update the form to ask:
- How often do you intend to visit? [constant|daily|weekly|occasionally|etc]
- How would you rate your skill/experience in <proposal> [beginner|intermediate|advanced|expert]
- How many questions OR answers would you commit to giving in the first 3 months? [0..20+]
Or whatever questions make sense - the idea being to make the commitment a positive thing, rather than a burden.
The answers can then be used to make a more informed decision (i.e. An expert who visits daily and will contribute lots of content should carry more weight than an occasional newbie) and the commitment percentage can be adapted to make use of this extra information.
Here's a mock-up for this idea:

It's not perfect, but it hopefully conveys the idea?
Other Stuff
I don't think it's important as suggested that people are deeply familiar with Stack Overflow - sure it needs some people that are, especially with regards to moderation, but most people (and especially so those that will commit to a Q&A site) are quick learners, and will have no problem picking up the ropes during the beta period, (under the guidance of the experienced SO/A51 users).
There doesn't need to be an overlap with Stack Overflow for a site to be succesful - again, as above, people can adapt pretty quickly, and tend to only need small amounts of guidance.
Specifically on the example proposal, 3D Graphic Artists are just as competent at learning websites as Photographers - it just so happens that photography is more common amongst SO/A51 users, and a popular activity in general, these days.
However, even if experienced StackExchange people are required, that is simply a matter of tweaking the existing commitment percent formula appropriately - but it doesn't mean it shouldn't be made easier for users to arrive from outside.