I do think there is something to this. Not just that they spend allot of money consistently but consistently stick with a project to see it to completion. In organizations that I have seen lack of progress they jump from one fire to the next. This means that they work on A for a few months and then suddenly drop it to work on B because B is higher priority. Only to later realize that A was more important and switch back. DARPA having a large budget also allows them to see through projects to the end.
DARPA contracts are structured with a lot of milestones and go/no-go decisions. We had teleconferences with the program manager and staff every two weeks and it was very clear that the money could abruptly stop if they ever felt that we were not going to deliver. This wasn't an idle threat, either: The program-wide PI meetings definitely got smaller and smaller as time passed. As a result, there was a lot of...reprioritization to make sure that the PM stayed happy.
This is very much unlike most other funding agencies. Once an NIH or NSF grant starts, you can be confident that the money will be there for the duration.
