Likely more than you ever wanted to know on the topic: https://benjaminreinhardt.com/wddw

A couple things I'd pull out:

Program managers sit limited terms to prevent empire building

Program managers have a great deal of autonomy once funding is initially allocated

DARPA has a high appetite for risk. They're ok with 90% of projects failing to hit their goal.

From your link:

> I would rather this be read by a few people motivated to take action than by a broad audience who will find it merely interesting. In that vein, if you find yourself wanting to share this on Twitter or Hacker News, consider instead sharing it with one or two friends who will take action on it. Thank you for indulging me!

You maybe shouldn't have shared that here, but thanks!

A focus.

DARPA research is performed in response to real challenges that service members and officers are faced with, and also in response to new capabilities that our adversaries develop.

Having done DARPA contracts, the difference I saw is that they feel more like moonshots. They are more fun to work on because you're encouraged to let your imagination run wild. Since failure is almost guaranteed it is kinda liberating since you don't feel the pressure to avoid failure. Failure IS the process here. Working on a DARPA contract is really the space where "there's no dumb ideas." Of course there actually are, but there's much more freedom to suggest these because at worse you rule some things out. And often the dumb ideas lead to good ideas that haven't been seen because the path starts at something that is clearly absurd. But then again, the premise of the problem is absurd so you have to try absurd things.