[WIP] explore reverse indexing in LCA v2 databases

[ctb]

Ref #306 #581 and #533.

In #533, we ensured that LCA databases contained all input hashes, whether or not the owning signatures had a lineage assignment. This makes them good candidates for general reverse indexing, but modifications are needed to make this functionality usable and friendly. Here is where we are experimenting with better functionality.

One question to think about in this PR is this: is sourmash lca about taxonomic stuff, or about all LCA databases, and if the latter, should we rename them from lca databases to something else, and/or should we support taxonomic information in regular signatures (which would support them in SBTs)? Deep thunks to be plumbed.

Specifically, this PR:

• adds a sourmash lca revindex command that mimics sourmash index but does not require a taxonomy spreadsheet;
• adjusts output of sourmash lca gather to be clearer when there is no available lineage assignment;
• tests behavior of sourmash lca gather when there isn't a lineage assignment!
• will maybe add phylogenetic summarization for lca gather as per Layer phylogenetic options on to sourmash lca gather? #583, since lca gather is now only a mild increment on top of gather (since [MRG] add lca DBs as inputs to 'sourmash search' and 'gather' #533);

Review checklist:

• Is it mergeable?
• make test Did it pass the tests?
• make coverage Is the new code covered?
• Did it change the command-line interface? Only additions are allowed
  without a major version increment. Changing file formats also requires a
  major version number increment.
• Was a spellchecker run on the source code and documentation after
  changes were made?

[ctb]

This PR is now superseded by #946 but I'm leaving it open until I think through the issues raised in the description, and maybe make them into new issues.