mpt: Unnecessary magic string in proofs

Currently all MPT inclusion and non-inclusion proofs start with the string "mptproof". Why is this necessary? [CT proofs](https://www.rfc-editor.org/rfc/rfc6962.html#section-2.1.1) don't have a magic string, and I don't know of any issues that people have run into about their non-self-describing-ness.

Concretely, the 8B are not likely to make a huge difference (though, at scale, maybe you could make that argument). But it just seems not necessary, so I figure I'd flag it anyway.

Start by locating the MPT inclusion and non-inclusion proof encoding and verification entry points, then read how the "mptproof" prefix is used. Compare that design with the linked RFC 6962 CT proof format and determine whether removing the prefix preserves unambiguous verification; done means reaching and documenting a clear decision.

Written by the indexing model from the issue text.