Breaking this out of #88 to make it separately decidable

Even a single "version" of SSVC like v2 has multiple trees -- does the vector represent all of them simultaneously? Or do we need to be able to specify which tree within a version this applies to? (e.g., SSVC/v2/coordinate/vul_id/key:value.../timestamp)

This may lead to an issue with tree versioning (and whether tree versioning is just the SSVC version or if there's a separate versioning scheme).

This may intersect somewhat with #109 in that if the decision here is to have a single vector represent all possible trees, it seems to imply that parameter omission in #109 should be permitted (otherwise vectors will be really long).

However, if the decision here is that vectors include a tree identity (the role idea in #106) then it seems like the omission/inclusion question in #109 is less relevant here.