If I was putting my own money in a project I'd definitely prefer the top 50% programmers with good organizational and interpersonal skills and concrete business plans over the top 5% developers who might just use their near-market salary to contribute to their favourite open source projects and work on esoteric problems that interest them. (I'd pay more attention to technical ability with "VC-istan" growth expectations and incentive structures)

I have a feeling that organizations part-funded by governments that hire intellectuals on near-market salaries to work on projects that interest them (and do the odd bit of teaching) already exist: they're called universities. I realise the plan here sheds a layer of bureaucracy and gives the productive a slice of potential profits they generate, but I think the comparison is reasonable, and I certainly wouldn't pump money into academia with the expectation of collecting returns on it.

Very good points. Thanks so much. I welcome disagreement because, without it, my ideas would stay in a raw crappy form. So... here are my thoughts.

My end goal is for this to congeal into something like an open-allocation environment, but one that exists across companies to some extent. I feel like if we bring OA to the external market, we'll force the hands of employers who will have to improve conditions if they want to compete for talent. So, the trans-corporate open allocation (because each person carries an income that follows her whereever she goes) is the ultimate end goal. That may be completely insane, but isn't that what the open-source gift economy is-- trans-corporate collaboration?

Getting designers, marketers and domain experts will be hard. I grant you that. Strong designers and marketers I can spot; domain expertise I would not be able to tell, unless I had it myself and, of course, for most domains I won't. I think that to get a domain expert, you'd have to hire her into a company already formed through this process, because you're right that there's no way we can build up a competent selection process for every domain we might care about.

I am not sure I get your second paragraph, but it seems to me that you want to build some kind of consulting company owned by top coders. Your offer would be to offer programming services, that only very competent programmers could do. My guess for such services would be: ultra scalable web apps, or creation of DSLs - it has to be something very advanced in order not to compete with the numerous consulting companies that are already existing, most of them doing mediocre stuff.

And I can see this taking off. Obviously not fast because most of your potential customers wouldn't even know yet they need your services, but that should be able to slowly bootstrap. And I suppose once you would have built a strong team and gotten some notoriety, you would be able to bootstrap your technological edge into some concrete product.

Having said that, I still have some doubts, mainly because I think we are in a coding bubble, and the future belongs less to the programmers than to the engineers who will build the new space rockets or the technology to exploit green energies from example, while with the standardization of functional programming and the development of various libraries it will be easy for a non programmer to start creating software components that are currently requiring programmers (IMHO there will be less "pure" programmers and more "X's who can program" as you once said here on HN).

Nevertheless I wish you can succeed with your idea, even though I don't think I'll be able to work in your company as I am not a top programmer.

I think you might be barking up the wrong tree here - OP's plan is to create an investment fund that focusses on long-term investment in talented people, rather than the VC approach of investing in "whatever facebook/twitter/google might buy to acquihire the founders".