Rethink how we abstract key management types

[ArPhil]

We should rethink how we "abstract" different key management types. Currently we differentiate on a very fine granular level (e.g. did methods). (How) can we abstract / group different key management types in order to provide more clarity? On a DID method level it is probably to much.