It's at least some comfort how far ARM has come the last few years, first Microsoft and soon enough Intel will be, still around and making money certainly, but decidedly behind the curve.
Next time I read about Richard Stallman's laptop I'll probably be scoffing a bit less.
And I think it has been shown that can happen. The most obvious example is the personal computer. For many years, computers were only available through large organizations. Eventually hobbyists found ways to make them affordable and distribute them to many homes.
I believe that the designs for some type of Intel processor and single board computer must be out there being used by quite a few people, say in certain areas of China. So they aren't really secret.
What we need is open source hardware, and personal fabrication systems.
Another example: move away from centralized ISPs to meshnets. If meshnets don't provide adequate latency etc., we just have to start coming up with our own alternatives to ISPs. Maybe just laying fiber strands on the street with some kind of glue or something crazy-sounding like that.
Re content: the loper-os.org guy is working on foss lisp-based hardware. Check it out.
The loper os thing is currently two guys playing with fpgas, but the one knows his lisp and knows what he wants and has supposedly written most of the OS code already, so I'm hopeful. I'm not sure if anything other than a networked lisp machine would work for the kind of verifiably-not-backdoored software/hardware that everyone wants. Aside from the obvious smell of a lisp hacker trying to do his own full stack, I'm surprised that it hasn't gotten more attention.