On (1) : absolutely, and I wouldn't want bitcoin to replace ECC with PQ algos. I would want PQ algos added. So, if I choose to transfer to an address that also uses a PQ algo, I would then need to sign with ECC *and* with that PQ algo. For big amounts, I wouldn't want to risk it being at the mercy of a technological advancement.

On (2) : higher fees : if bitcoin Core wants to consider that, there is a lot of work to do in making it harder for spam data to price out actual monetary transactions. The higher fees argument might have hit harder last year but v30 changed that (on top of what taproot and segwit did). They even changed the definition of bitcoin to make it into a jpg network instead of p2p money.

On (3) : on killing cheap nodes : again, same as with (2), v30 pretty much affirms that Core doesn't care about that. They'll have us relay whatever spam spammers want to spam us with. Pruning is not an interesting option because it kills archival nodes and kills electrum servers.

If we fix the bad incentives to spam the network with non-monetary transactions (thereby limiting the undue competition for blockspace), we can afford to protect the network by giving the option (for those who prefer that) to also have a PQ algo in addition to ECC.

That might actually make blockspace more competitive, while keep bitcoin a p2p money network, and might be good long-term for the network when the mining needs to run only on miner fees.

Many serious people are saying that quantum will be a threat. I'm not technically competent enough to know one way or the other. But giving what's at stake, adding an optional (additional) PQ seems to be the way to go.