This is an interesting concept. I think the implementation would be very tricky though. The most important implementation detail is whether the investments are chosen by the investors, or by a manager or management committee of some kind.
I think you would have problems doing this with investors doing the decision making; I would imagine you'd need to spend at least 10 hours per company invested to make a remotely informed decision; people aren't going to want to spend hours making a decision about a $50 investment. If you get the time commitment much lower, it will probably be a huge magnet for fraud.
With a manager-based decision making process, I don't think the JOBS act affects the legality much. You'd probably need to house the decision-making in a registered investment company, which will probably cost you at least $50k/year in compliance costs. Getting even 1 remotely qualified manager would probably cost you $200k/year. I don't know that people are going to be as excited about the idea when they hear the expense ratio is 25%/year.
To make it work, it seems to me you would need a national-level parent organization to amortize compliance and legal costs, and (qualified) volunteer decision makers.
If you look closely, my numbers don't quite add up -- $50/mo is $600/yr not $500/yr.
I was guesstimating about $200k/yr in operating expenses (and that's with a lot of volunteered expertise, as you note).
But structured as a social-purpose non-profit, with everyone paying in being a member who elects a management board, who allocates the money as they see fit (perhaps with some constraints chosen by members), I think this actually bypasses a large amount of compliance work -- if you never return the money, and are strictly a non-profit investing in local businesses which rolls returns in to future investments -- because the members aren't actually making investments: they're paying for a non-profit's operational funds to support local business ventures. Structuring this way also makes for a better case to get donated expertise, eg from local lawyers and investors, and probably makes your tax situation a lot better.
You probably need some rules about earning lifetime membership after a certain pay-in, and definitely some about conflicts of interest (eg, funding member's ventures) but those are more technical details.
It's not a good financial investment vehicle, in that even in the best case you'd almost certainly have negative returns and in the worst, you'd have no returns at all -- but if your goal is changing the function of the market to support local enterprises and you're willing to pay in to make that happen, I think it'd function excellently. (Of course, you only really need to seed one or two Starbucks to have the whole portfolio right-side up again, even after decades of total losses. So at a generational level, it might work as a financial instrument as well.)
Worth noting, even as a non-profit, there are ways to "return" the money besides the value of the businesses started if you do find yourself with extra funds -- investments in local parks, schools, art programs, etc. It'd never come back in currency returns, but I expect "investors" would see tangible value in their lives, quite possibly exceeding the monetary value they put in.
tl;dr: It's easier if you structure as a one-way investment in your community, and collect dividends from the output of a vibrant community.
I would prefer a for-profit one, but due to many of the concerns you point out about compliance (and that you can no longer get donated labor), I don't think you can make it work (you end up spending too much on compliance and management relative to the size of the fund). Amortizing the costs across a national fund substantially changes the incentives and control structure -- which replicates many of the problems I have with current investment mechanisms. (And you'd still run in to issues where every state has special snowflake rules about certain things.)
The reason is I think you really need to focus on localization of such funds to make them effective -- metro scale ones. You can probably make the financials work by simply eating the overhead on that scale, but not on one trying to turn a profit, but paying for labor. Losing 17% of your investment every year to overhead is simply too rough, because you'd need to generate 20% returns.
You could do it, but it would substantially change the nature of the investments you could make and the size the fund would have to be to operate. There's a certain charm to funds in the 1-10 million dollar range making small investments in 10-100 businesses. (And I believe helps combat some of the problems involved with corporations, by intentionally distorting the market by "losing" money purposefully and recouping the value from the social change.)
I was (obviously) unaware of CDVCA, but even there, it seems like they're mostly a broader investment scale and set up as a traditional venture fund (ie, looking to connect traditional investors with high growth potential ventures) as opposed to operating a continual flux of small, local businesses at (likely) negative returns funded by the local populace at large. (I poked around the NYC one, for instance.)
That said, they likely have worked out some of the legal hurdles involved.
I think you would have problems doing this with investors doing the decision making; I would imagine you'd need to spend at least 10 hours per company invested to make a remotely informed decision; people aren't going to want to spend hours making a decision about a $50 investment. If you get the time commitment much lower, it will probably be a huge magnet for fraud.
With a manager-based decision making process, I don't think the JOBS act affects the legality much. You'd probably need to house the decision-making in a registered investment company, which will probably cost you at least $50k/year in compliance costs. Getting even 1 remotely qualified manager would probably cost you $200k/year. I don't know that people are going to be as excited about the idea when they hear the expense ratio is 25%/year.
To make it work, it seems to me you would need a national-level parent organization to amortize compliance and legal costs, and (qualified) volunteer decision makers.