I kind of wonder if they're fishing for the next batch of project managers for information systems projects.
The previous director of DARPA, wasn't so popular with the research community; he was an engineer from industry, and instituted a regime of (fairly unrealistic) GNG (go/no go) targets in programs every year. If sites didn't hit certain scores on the GNG evals, they lost their funding. Which sounds not too bad on the surface - why would you continue to fund an organization that's not doing well? Except that for a number of projects that DARPA does (information and language processing technology, in particular), there's some very difficult problems with have somewhat good problems where you might be able to shave a few points off with refinements of existing approaches. If your funding hinges on meeting targets, rather than trying to solve the problem, you're just not going to pursue a new (but possibly fruitless) approach because the risks are too high.
I kinda viewed that wording as looking for new, fresh, research oriented blood for the organization; perhaps there's an ongoing return to the roots of DARPA - research. Research with a purpose, directed towards a specific goal, with producing something useful at the end, but research that may fail but needs to be done to discover possibly novel approaches to problems that might otherwise go ignored in a wholly results driven system. One can hope.