docs: Book Comparison wired to the prop card (matrix row 18 DONE, 16/26)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VsztNChZ7vEvSR61AuMhD1

specs/STATE.md

@@ -74,6 +74,17 @@
 > confirming audit still needs weeks of settled lock+close+outcome rows.** Volume ~1.5-3k
 > rows/day. Populates on the next snapshot tick (14:00 UTC); no promotion, no served edge,
 > no capture-of-close change. Next: let it accrue, then re-run the staleness audit.
+
+> ## 📚 BOOK COMPARISON WIRED 2026-07-29 (commit 2ab2eea): matrix row 18 → DONE (16/26)
+> The dead `BookComparison.tsx` is routed to the GradeResultCard via a new self-fetching
+> `BookComparisonPanel` (reads live `/api/books` = fenced `bookprices`). Contract fix: books
+> frequently sit at DIFFERENT lines (WNBA DK 21.5/FD 18.5), so it renders EACH book's own
+> line per-row (no false shared-line comparison). Honest states: single-book → one book +
+> "One book posting this prop.", no crown; multi-book → all books, none crowned
+> (`BOOK_CROWN_ENABLED=false`); no books → NULL (self-hide). Display-only: only the
+> always-empty inline `d.books` section was replaced — grade/projection/line/PriceTriplet
+> untouched (web-only, grade byte-identical trivially). Freshness equal (same snapshot as
+> the lock; intraday touches neither). HELD: push-to-book, movement strip, crown enable.
 - **Redirect EXISTS + WIRED:** `closingCapture.buildCaptureRows`→`closing_captures` (append-only,
   provenance: captured_at/book/line_type/both-prices/missed_reason) via `intradayRefreshService:221`
   + internal endpoint; `ledgerService.attachClosingProb`→`closing_prob` (de-vigs both raw sides,