>
If I find out I'm in the economic minority then I have little choice but to either accept the existence of the new rules or sell my Bitcoin
I do worry about what I have called a "dumb majority soft fork". This is where, say, mainstream adoption has happened, some crisis of some magnitude happens that convinces a lot of people something needs to change now. Let's say it's another congestion period where fees spike for months. Getting into and out of lighting is hard and maybe even the security of lightning's security model is called into question because it would either take too long to get a transaction on chain or be too expensive. Panicy people might once again think something like "let's increase the block size to 1GB, then we'll never have this problem again". This could happen in a segwit-like soft fork.
In a future where Bitcoin is the dominant world currency, it might not be unrealistic to imagine that an economic majority might not understand why such a thing would be so dangerous, or think the risk is low enough to be worth it. At that point, we in the economic minority would need a plan to hard fork away. One wouldn't necessarily need to sell all their majority fork Bitcoin, but they could.
That minority fork would of course need some mining power. How much? I don't know, but we should think about how small of a minority chain we could imagine might be worth saving. Is 5% enough? 1%? How long would the chain stall if hash power dropped to 1%?
TBH I give the world a ~50% chance that something like this happens in the next 100 years. Maybe Bitcoin will ossify and we'll lose all the people that had deep knowledge on these kinds of things because almost no one's actively working on it. Maybe the crisis will be too much for people to remain rational and think long term. Who knows? But I think that at some point it will become dangerous if there isn't a well discussed well vetted plan for what to do in such a scenario. Maybe we can think about that 10 years from now, but we probably shouldn't wait much longer than that. And maybe it's as simple as: tweak the difficulty recalculation and then just release a soft fork aware Bitcoin version that rejects the new rules or rejects a specific existing post-soft-fork block. Would it be that simple?