I used to be keen on what number of potential combos of BIP 32 derivation paths exist, which might lead to a public key derived from a seed. Whereas some folks colloquially name it “limitless,” I do not assume that is technically true.
I imagine the maths then is (2^32)^255 which is 2^8160 distinctive paths.
Whereas the variety of potential public keys is way smaller at solely 2^256.
Subsequently, it appears cheap to conclude that every seed, utilizing some unknown derivation, can produce each single potential public key (many instances over, with collisions).
Clearly this is not a sensible concern, however it might be fascinating from a trivia perspective. Can somebody affirm or level out a flaw in my logic? Thanks!