docs(backlog): defer §6 CritiqueFinding/calibration until a consumer exists

BACKLOG.md

@@ -56,10 +56,15 @@ and §5 reviewer prompt/needs wiring (3467826) are shipped — RETRO candidates.
       (resolved upstream) — no output-parsing, no shared-findings schema needed. `ProcessStaticAnalysis-
       Runner` is the injected adapter (nullable seam; null → logged no-op). Prompt half was `3467826`.
       QA gate in §F → `docs/qa/QA-static-first-reviewer.md`.
-- [ ] **§6 `CritiqueFinding` schema in code** — only test expectations exist today; no
-      structured type. Shared by plan critic + code reviewer.
-- [ ] **§6 `CriticCalibrationProjection`** (keyed by modelHash, role) — needs runtime
-      history to exist first (spec ordering #5). Same projection as pipeline-addenda A3.
+- ⏸ **§6 `CritiqueFinding` schema + `CriticCalibrationProjection` (= addenda A3)** — **DEFERRED until
+      calibration is actually built.** Substrate with no live consumer today (assessed 2026-06-14): zero
+      `CritiqueFinding` references anywhere (the old "test expectations exist" note was stale — none do),
+      no `CriticCalibrationProjection`/`CritiqueOutcomeCorrelatedEvent`, and **no plan critic** — so the
+      spec's "single definition, two consumers" rationale has zero consumers. The one reviewer that exists
+      emits `review_report` = `verdict` + free-text `notes`; code reads only `verdict`, and the implementer
+      reads `notes` as prose fine. Structured findings only earn their place when the calibration
+      projection is built — which needs runtime critique/outcome history first (spec ordering #5) and a
+      second consumer (a plan critic) to justify sharing. Build the schema *with* calibration, not before.
 - [ ] **§4 architect contradiction-check** — L3 retrieval over decision journal + ADRs →
       `PossibleContradictionFlag` (display-only in v1). Unbuilt.
 - [ ] **§3 symbol grounding** — deferred; needs a real symbol index (file-path grounding done).