It would be useful to:

(1) be able to distinguish what data - POSIX or SID - were requested/returned.

Currently it is not possible as cache request being used are the same.

For example both:
nss_cmd_getsidbyuid() -> nss_getby_id(CACHE_REQ_USER_BY_ID)
and
nss_cmd_getpwuid() -> nss_getby_id(CACHE_REQ_USER_BY_ID)

results in something like:

[nss_getby_id] (0x0400): Input ID: 822000000
[cache_req_set_plugin] (0x2000): CR #11729533: Setting "User by ID" plugin
[cache_req_send] (0x0400): CR #11729533: New request 'User by ID'
...
[cache_req_search_send] (0x0400): CR #11729533: Returning [UID:822000000@some.domain] from cache

Important difference is that sid_by_* requests are not served by mem-cache by design. Thus it is handy to be able to differentiate the type of request in a log.

(2) be able to match client ([get_client_cred] (0x4000): Client creds: euid[0] egid[0] pid[3915]) and request (CR #11729605: New request)
This would be useful to identify source of request flood.