That strikes me as a dangerous path forward.
I don't actually think there is anything wrong with this: "everybody eventually gets tired of arguing angels-dancing-on-the-head-of-a-pin, and we're left with the status quo"
What gives Bitcoin value aren't its technical merits but the fact that people believe in it. The biggest risk here isn't that 20MB blocks will be bad or that 1MB blocks will be bad, but that by forcing a hard fork that isn't nearly universally agreed upon, we will be damaging that belief. If I strongly believed some hard fork would be better for Bitcoin, say permanent inflation of 1% a year to fund mining, and I managed to convince 80% of users, miners, businesses and developers to go along with me, I would still vote against doing it. Because that's not nearly universal agreement, and it changes what people chose to believe in without their consent. Forks should be hard, very hard. And both sides should recognize that belief in the value of Bitcoin might be a fragile thing. I'd argue that if we didn't force through a 20MB fork now, and we ran into major network difficulties a year from now and had no other technical solutions, that maybe we would get nearly universal agreement, and the businesses and users that were driven away by the unusable system would be a short term loss in value considerably smaller than the impairment we risk by forcing a change.