i'm not sure what you mean exactly.
Here is some example, where you have 1-of-2 multisig, without using P2SH:
https://mempool.space/tx/23b397edccd3740a74adb603c9756370fafcde9bcc4483eb271ecad09a94dd63it can go up to 15 compressed keys apparently. not just 3.
https://github.com/bitcoin/bips/blob/master/bip-0011.mediawiki#specificationSpecification
A new standard transaction type (scriptPubKey) that is relayed by clients and included in mined blocks:
m {pubkey}...{pubkey} n OP_CHECKMULTISIG
But only for n less than or equal to 3.
Of course, you can get more than three keys, but then it is non-standard, if you use bare multisig.
so first of all why would parties use keys from the same deterministic wallet as private keys for their multisig wallet. that wouldn't really make much sense.
Because there is a difference between regular usage, where public keys are safe to use, and some attack, where ECDLP is solved. Because if single public keys would become unsafe, then attacking N keys simultaneously wouldn't require much more effort.
So, if ECDSA is safe, then you can use 2-of-2 multisig, and be sure, that both keys are always needed. But if ECDSA is no longer safe, then finding a distance between the generator and the first key, is not that much harder, than finding a distance between all of those three points. Because eventual attacks are as complex for two points, as they are for N points, it is all based on finding collisions and relations between keys, no matter, how many of them you start with.