Names the log door the answer-parent entry stamps

[johnnyt]

Bead: sp-f1s (filed by the campaign-034 conductor from sp-t12's two open
items; scheduled as an SF035 lane P fill).

What changed

Two things, both about the same stale sentence.

1. lib/statifier_persistence/driver.ex - the comment above
   respond_to_parent/3 said entry: :answer_parent "is telemetry only".
   ADR-0010 decision 5 made that false: on an adapter that keeps an input
   log, entry: also stamps the door of the entry appended to the
   parent's log, through the done_invocation/5 / failed_invocation/5
   re-entry that function performs (decision 5's table gives
   :answer_parent an "appends: yes" row). The comment now names the log
   door. invoke_id: and child_count: keep the telemetry-only claim,
   which is still true of them, and lose the "too" that tied them to a
   door that no longer carries it. runs.ex's t:opt/0 doc already says
   this correctly; only driver.ex was left behind.

2. docs/adr/0010-durable-per-run-input-log.md - a dated Note under
   decision 2 marking its "append_input, list_inputs, input_log,
   supports_input_log? and Recording currently have zero occurrences in
   lib/" as the authoring-time ground for the naming choice rather than a
   standing claim. sp-80g landed the four callback and facade
   names, so the sentence reads as falsified without the Note; it was never
   a decision, only the reasoning for one. Recording is the exception and
   stays upstream's - decision 8's mapping onto it is named in the record
   and built nowhere - and the Note says so (cure for the pass-1 direction
   review finding, commit 12289f9). Amend-by-addition in the record's
   own established Note shape
   (the record already carries ten dated **Note (...)** sections,
   decision 1's at lines 139-150 among them). Zero removed lines in docs/adr/.

Acceptance criteria

• "The comment names the log door" - checked: driver.ex:794-801 now
  names door, the parent's input log, and ADR-0010 decision 5.
• "the Note if taken adds zero removed lines" - checked:
  git diff origin/main -- docs/adr/ | grep -c '^-[^-]' returns 0.

Provenance

• File map as dispatched: lib/statifier_persistence/driver.ex (the comment
  lines) and docs/adr/0010-*.md (the optional Note). Nothing outside it was
  touched.
• The dispatch brief located the comment at driver.ex:717; on main at
  19f2ced it is at :794. Reported as a brief correction, not acted on
  differently.
• No changelog fragment: per changelog.d/README.md this is invisible to a
  caller of the public API (a comment and a record Note).
• Runs beside sp-e8c (storage.ex moduledoc); the two touch disjoint
  files. No rebase-over-sibling resolution was needed.

Verification

Full mix quality green on the branch head, attested via
mix quality.verify (attested: true, scope all, no profile): format,
compile, credo, dialyzer, deps, 482/482 tests, 95.7% coverage. Postgres was
the local server on localhost:5432.