dropck: uses of terms "superregion" and "subregion" suboptimal

[pnkfelix]

Spawned off of: #21022 (comment) , #21972

The dropck code (RFC 769) added a new SubregionOrigin variant for lifetime-inference errors, but the error message it reports is probably not written in terms we want to expose to our end users.

(However, this may be a moot issue since pnkfelix has most often seen the lifetime constraints added by dropck as leading to other lifetime-inference errors, and so he does not actually have a test case available at the moment that illustrates this issue.)

[pnkfelix]

cc #22321

[steveklabnik]

Triage: over a year later, do you still feel this might be moot, @pnkfelix ? or a test case?

[pnkfelix]

I continue to see no instances of this diagnostic arising.

I'm thinking I might close this issue, though I might first want to confirm that the code in question is going to be completely obsoleted by NLL anyway.