perf: combine whole fine storage and batched stream completion

Integrate only the two reviewed improvements onto eec4677. Storage tiles use
one whole wide fine allocation; step(n) completes its unchanged one-stream
schedule with one wait per call. Raw exchange retains the final interval;
an open exchange window spans the completed batch.

Storage-source: ab699854ae2fb0c994a14bfd7124dddbe50d8a88
Original-sync-author: Claude <claudeworker@noreply.invalid>
Sync-source: ad1d82cb7f
Sync-doc-correction: 512dcc923b
Validated-local-sync-source: 65c001c3294a6eb45283568bf4967c673b33b19a
Sync-stable-patch-id: a825dcd47b33c5c4d9be31e0071d2306eb344d0b
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

docs/validation/refined_dfg_window.md

@@ -56,3 +56,16 @@ Root must independently review and admit any execution. An admitted paired pilot
 the existing watchdog, device selection, `--gpu-gates-approved`, positive
 `--max-wall-seconds` and a new output directory, adding only `--exchange-window` to the
 selected collision's pilot command. No GPU or remote action was performed for this patch.
+
+## Wide fine storage allocation
+
+The wide runner uses `FINE_BLOCK = (240, 568)`: one block containing the whole
+fine allocation, including the four-cell refinement ghost band on each side.
+The active `232 x 560` region alone is not a valid substitute. Physical bounds,
+resolution, numerical kernels, 256-thread launches, per-root synchronization
+and all scientific gates are unchanged. The plan records this storage block.
+
+This is a wide-only configuration backport; no compact-placement or public-Jet
+API is added. Existing source hashes remain checked, and the separate runner
+hash requires a fresh completed pilot before a full run. Earlier short-layout
+timings do not establish a full-run speedup or change historical failed verdicts.