Please kindly make substantive arguments, so I am not obligated to pollute my Comments list with moar replies. I am obligated to reply now to your (more or less vacuous) comment, because otherwise naive readers might incorrectly assume you’ve stated some substantive objection. Please don’t take this as a personal insult. It’s just that instead of editing your comments to merge new comments, you make double replies for example. I’ve written too many comment replies today. Please in the future consider to merging your new thoughts by editing a previous comment. I will notice it and edit my reply as well to add any reactions. No need to pollute my Comments list such that I will not even been able to find anything anymore because buried under 3000 comments.

One other potential thing to consider is the bugs that could result from such a rollback after hardware changes. I don't think there would be a problem with ASIC's yet since they have one thing to do which is find the nonce, but the longer segwit remains as a standard, the more likely other hardware changes will make it increasingly too difficult to roll back.

You have not proposed any such specific hardware dependency and inertia.

Only hashrate matters.

The only plausible SegWit hardware dependency I’m aware is Gregory Maxwell’s changes to defeat surreptitious (i.e. undetectable) use of ASICBOOST. Yet I see new ASICs being announced with ASICBOOST. So apparently detection is not a sufficient deterrent.

Mining is not really anything like a consumer OS such a Windows. The analogy is weak. You should have tried to find a substantive example first of your hypothetical. I believe there is none to be found.

What's the chance that a fork of Core from the bitcoin classic wouldn't be like Eth and Eth classic with Core being like Eth in marketcap? IOW, what other violations of economic principle have Core engaged in?

There was a time when I was thinking about all that deeply and might have had some ideas to share, but I let it go. I have to focus elsewhere.