Very interesting thread.
My background is 20 years of designing CPUs/ASICs/GPUs.
A few comments here:
1) There is no computational problem that you can't design custom ASIC hardware to do faster than a GPU.
Can we make a proof of work based on the mathematical principles used for rendering video games? GPUs should be pre-optimized to this task.
Half joking. But seriously?
x86/x64 CPUs are too unspecialised to have an algorithm made for them i would assume