No. I think the reasons that you want to have decentralization in development are well laid out in James Scott's book, _Seeing Like a State_. Speaking broadly, state bureaucracies produce one-size-fits-all, middle-of-the-road solutions. They aren't very good at shipping products rapidly. They aren't very good at making risky long-run bets. They aren't very good at tolerating conflicting opinions.
Decentralized development processes--which seems here-and-now to mean "firm," "market," and "patent," since we don't know how to do them any other way--are relatively good at making risky long-run bets and at shipping products rapidly. A diversity of organizations ensures at least some diversity of opinion. But the big problem with "firm," "market," and "patent" is that the stuff produced is someone's private property, which they then sell at a monopoly price.
So at the moment we are stuck: we (mostly) do basic research at the public level, believing that the benefits of widespread free distribution outweigh the costs of bureaucratic centralization; and we (mostly) do applied development at the firm level, believing that the benefits of entrepreneurship, enterprise, and "hard incentives" outweigh the costs of granting the drug company a temporary--temporary--monopoly over the products developed.
But how about other institutional forms? Things that are neither the bureacratic state or the market with entrenched intellectual property rights? One might think--indeed, I know a president or two of a highly-endowed private university who does suspect--that a university whose mission is to spread knowledge and that is secure enough to allow its researchers to be free from bureaucratic fetters might provide a superior mode of organization (as long as its deans can be kept from focusing on how much money their schools might make if they sold rather than gave away their knowledge). But it is not clear to me how one organizes a semi-non-profit in such a way that it attains the benefits rather than the drawbacks of "state" and "market"...
Brad DeLong