This is how DARPA funding typically works. Look for proposals on some crazy idea that is way beyond current state-of-the-art, seed money if anything halfway reasonable comes in, and then hope for the best. Much more so than any other large research body (in the US at least).
There was a DARPA program manager quoted a few years ago in the MIT Tech Review saying they try not to have a greater than 15% overall success rate on DARPA funded projects...any more and they are not dealing with hard enough problems.
― Silas Beauford (Silas Beauford), Saturday, 17 April 2004 00:42 (twenty-two years ago)