Loaded from a snapshot; a change journal reported nothing touching this subtree
since the cursor, and nothing has re-checked it.
Named for what actually happened — a journal scoped the work — rather than for
what a reader might wish it meant. Nothing here was confirmed against the
filesystem: a scoped revalidation stats the paths the journal names and does not
stat the rest, so this value rests on the journal having been complete.
It is deliberately weaker than Self::Revalidated because that assumption is
known to fail. macOS FSEvents will report HistoryDone after silently dropping
history, with no degradation flag, which means a journal answer can be wrong
without announcing it. Journal-assisted revalidation therefore bounds exposure
with a maximum age and a periodic full sweep; those are risk controls, not
proofs, and they do not make any individual answer here verified.