Bitcoin is designed so that different asymmetric crypto governing controlling coins can be compatibility introduced-- as has been done a couple times in the past (to introduce revisions to the script system). PQ crypto could be introduced in that way.
Unfortunately existing PQ signature schemes don't have a great mix of maturity and performance that make them especially attractive absent a clear and present need.
For years I maintained a private patch set to introduce hash based signatures on short notice if needed. I haven't for a number of years now in part because there are now enough contributors I'm confident that if it were needed it could be done quickly without any special preparatory effort.
Users can change the rules governing their coins just by moving them.
We know how to construct kinds of forward compatibility that could allow being spent with a new scheme but the specific scheme would have to be set in advance so they users could generate the future keys at the same time the generate the current ones. And, as mentioned, the existing choices aren't that great (for this application).
If Satoshi Nakamoto did lose his keys to those treasure trove wallets, those will be up for grabs, as will anyone who lost keys.
Perhaps hoarded encrypted PCAP's aren't the only motivator for quantum opportunists.
What's the technical feasibility of migrating BTC's pow scheme to quantum-resistant hashing?
What is the technical feasibility of migrating wallet/private keys to quantum-resistant crypto?
Would any such migration of wallet/private keys be dependent on public/non-public interactions with the blockchain by wallet holders, and would those interactions introduce deanonymization or other risks?