Any seven page treatise on big blocks that doesn’t have the word “quadratic” in it, or “denial of service”, or “incentives” is not worth reading.
Larger blocks need to be engineered… not thoughtlessly shoved down everyone’s throat. There are solutions out there… incentive structures to encourage node and network growth, compact block relays to reduce the latency impacts of larger blocks. But damn, you just gloss over the fact that it’s actually hard to get this stuff done right.
Also the HK agreement was “segwit first”. Core devs have BIP103, and there’s a BIP100 as well, and there’s a number of pull reqs for hard forks that are much safer than just “raising limits and seeing what happens”. So progress is being made on safely increasing non-witness limits.
But you pretend that this isn’t the case.
Why? Why write a huge essay on something and ignore the real research and effort being made to get the job done?