I agree almost with all you wrote, but it has to be noted that the problem with Bitcoin is slightly more complex due to the reaction time of the system being slower.

Banks and their software providers can update their account systems to post-quantum cryptography in a couple of days (most smaller banks at least are anyway using cloud solutions today which are even faster to update, perhaps in hours...).

But in the case of Bitcoin, the biggest issue is probably that while you could also deploy a softfork relatively fast, if there are too many people which have still vulnerable keys (either P2PK/P2MS etc. or re-used addresses) then that would put a lot of pressure to the block space market, potentially for months or even years.

That's the reason why I think people should begin to care about at least not re-using cold storage addresses, which removes the "initial threat" already -- coins which can be attacked with years of quantum computing work. And that the devs should come up with a post-quantum solution before 2030 or so. So the migration can be gradual and slowly without any interruption of Bitcoin's functions.