I honestly think it's much easier to just convince I3 to give you guys a bunch of money to spend at your discretion. I would only devote precious DAC development resources to actual profitable DACs. Definitely a good idea in theory though.
If centralized job or bounty auctions are profitable, and centralized exchanges are profitable, why wouldn't it be profitable to build a decentralized job/bounty auction site and have a decentralized exchange?
Also once again why should the community exclusively rely on I3 to give us a bunch of money when we can just develop a DAC to decentralize the process of resource allocation? We can come up with an algorithm that we agree upon by consensus to be the most fair and then let the market go to work.
If people think shares in a DAC which handles resource allocation and bounties would be valuable why not develop it? I do agree it would take a lot of time and there are probably many other more important DACs to build first but I think if we are going the route of streamlined decentralization why shouldn't we go all the way?
If people in the community want to automate the process of crowd funding and allocating resources as much as possible why wouldn't that agenda be considered noble? It would make / allow DACs to organically build themselves from idea to finished product with an algorithm sending out shares according to Proof of Contribution. It of course would not be 100% autonomous because human beings would be working for the DAC doing things that cannot be automated but resource allocation is something which can be automated and Bitcoin proves that.
I think in the long term it's worth it to build it. In 5 years someone could come along, post an Angel address, people send money to it, and the DAC would take it from there. The autonomous agent would collect the funds and then start distributing it according to a pre-defined algorithm because certain tasks have to be repeated for every launch.
Make a power point presentation/whitepaper.
Make a main website.
Make a Facebook page.
Put website in signatures.
Create wiki.
Create articles promoting the DAC.
Debug and test the code.
Make blockexplorer.
All of the grass roots stuff could be simplified down to a script, and for humans a readable set of instructions which if followed step by step will create a fully functioning DAC. All of these tasks could be incentivized by distribution algorithm.