Refactor: Improve clarity of note syntax

[LHerskind]

The life-cycle of notes is currently fairly hard to follow. This is partly due to its creation before traits was available, and partly from the fact that parts of the nullifier computation depends on what the nullifier is to be used for.

E.g., when computing a nullifier to emit, that is different than when you wish to prove that some nullifier is in the state.

The difference is an effect of the insertion relying on the kernel to apply transformations before actually being inserted into state, and for reads it needs to apply those transformations itself.

It means that there currently are execution flows that "go in circles" which can be quite strange to follow.

[LHerskind]

#4207 reduces the entry-points for the note/nullifier slightly.

[LHerskind]

Discussions with @sirasistant and others lead to updating the compute_note_hash function name to compute_note_content_hash to avoid confusion with the actual note hash. See #4342

[LHerskind]

See #4344 for more changes related to the lifecycle naming.

[LHerskind]

#4383 updated the NoteInterface to handle serialization and deserialization specific to notes.

I'm not sure how much more we can change around the note syntax, unless we try removing the header entirely from the note, and have something like a WrappedNote as have been alluded to in other discussions.

[LHerskind]

For above mentioned reasons. I will close this issue, and create #4419 where this can be discussed.