This draft paper by Henry Farrell and Cosma Rohilla Shalizi looks like an absolutely fascinating project. Basically, it argues that in certain circumstances, democratic deicsion-making is a superior way of solving large-scale problems than its two primary competitors (markets and hierarchies). The argument builds on research exploring the importance of diversity in decision-making groups (one interesting paper they rely on is Lu Hong & Scott E. Page, Groups of diverse problem solvers can outperform groups of high-ability problem solvers, 101 Proceedings of the National Academy of Sciences 16385 (2004)), including the provocative claim that "diversity trumps ability".
In brief (and probably missing a ton of subtlety), all of us come at problems with different perspectives, which channel our ability to spot certain solutions to problems while missing others. People who are very much alike will tend to see similar solutions. Diverse groups see a greater breadth of solutions and thus are more likely to find the optimal one. Hong and Page present a model demonstrating that, for large groups in which there is a distribution of ability but all members are competent, it is better for decisions to be made by a random sample X of group members than by the X top performers (hence, diversity trumps ability -- or as William Buckley might put it, a society is better governed by the first 400 people in the Boston phone book than by the Harvard faculty).
Farrell and Shalizi argue that democracy simulates these conditions by putting decision-making in the hands of a broad, diverse cross-section of the community who are situated in a state of at least partial deliberation with one another. While this deliberation is not going to be at the idealistic level hoped for by deliberative democrats, it is enough such that the decision-makers will be aware of proposals and solutions made by others and can selectively (or even surreptitiously) incorporate them into their own plans.
I'm not doing the project justice, though, and I highly encourage you to read the linked draft. It looks like the start of a superb piece of work.