Wouldn't surprise me if high waste ratio but let's give credit where due. The first, open-source CPU/SOC's were result of ESA:

http://ramp.eecs.berkeley.edu/Publications/LEON3%20SPARC%20P...

Genode OS and some static analysis tools came up under FP7. I bet I'd find more with time.

That's my experience working closer to research teams in general. Mostly papers to be published, long philosophical meetings and very really code.

Not saykng it's bad, obviously some minds are fit for that. But it might contribute more to burn out if seeing something in action is what you're really in need of (going back to OP's question).

Can't agree with you more. I'm currently in the business with FP7 and Horizon. What a pile of... It's all about buzzwords and overly complicated "architectures" which should be as general and vague as possible to dazzle the reviewers.

That's actually the impression I get at my company too.