Why isn't a random salt used per credential for passkey-based encryption?

Unless I am reading the code incorrectly, it appears that the static salt "passwordless-login" is used for every credential for every user. According to WebAuthn Level 3, “If the evaluation inputs are unpredictable then even an attacker who could satisfy user verification, and who had time-limited access to the authenticator, could not learn the encryption key without also knowing the correct PRF input”. Additionally CTAP 2.2 requires HMACGetSecretInput to contain “32-byte random data” for salt1 and salt2 when using the hmac-secret extension (which PRF sit on top of).

Based on this information, wouldn’t it be better if a random salt were created per credential per user? All the server would need to do is store these salts to pass back to the user.

I can only post two links in one post since I am a new user, so here is the link to the CTAP 2.2 section.

Associating a salt per Credential ID requires knowing the user beforehand to obtain the Credential ID and associated salt; however when using passkeys—more explicitly userless login—this is clearly not possible. Put differently, when using passkeys (instead of multi-factor flows where WebAuthn is used as the second factor), this is only possible if logging in were a separate step from vault decryption.