Ok, I get it. But I still think this can be true in 2 of 2 multisig if only both public keys have not been exposed yet. But if it has been exposed, don't you think an attacker can brute force the public keys independently and maybe in parallel ( that is at the same time with two different quantum computers) to derive the private keys. [...] Yeah, your suggestion provides additional layer of security because it will cost more resources to execute but I think the attacker can spend such resources provided that the amount of coins at stake is worth it.
Quantum computing power will be incredibly expensive during the first years it will be available, and it will also take weeks or months to derive a single key. In this case, if you have two addresses with 1 BTC on them, and one is a multisig address, then the hacker will for sure first try the single address.
Once this "first post Q-day phase" is over and quantum hacking has been confirmed, I guess nobody will re-use an address anymore and post-quantum cryptography will be implemented. So this is a mechanism to increase the likelihood that you will not be a target in the first post-Q-day phase.
Come on, you can't really say this. What about the possibility that your single-sig entropy was not generated randomly, as with the Coldcard case? Or what if your passphrase is not secure enough, and can be brute forced by someone finding your seed phrase? Or what about the idea that a strong passphrase is not human readable, and you can make a mistake by choosing to write it down? (As opposed to another seed phrase.)
I don't understand the point of these examples ... these things can happen with any kind of key, not only with multisig, I guess the probability to mess this up is the same.
If you're worried about a key generation algorithm (?) you could generate two HD seeds and then use one of the derived addresses each.
(BTW, I meant passphrase in the sense of "seed phrase" and not some random "password" or so. I'll change this in the other post to the correct wording.)