Re: Naming of the different stats systems / "stats collector"

I'm growing fond of "cumulative".  It is more precise (and restrictive) than "metric" but that is beneficial here so long as it is indeed true (which a quick skim of Table 28.2. Collected Statistics Views [1] leads me to believe it is).

I'd be concerned that subsystem implies a collection process in a manner similar to how you associated datastore with a physical file on disk.  But I'd pick subsystem over datastore here in any case.

David J.

[1] https://www.postgresql.org/docs/current/monitoring-stats.html