preserve BLE disconnect race stress evidence

tests/fixtures/README.md

@@ -68,6 +68,16 @@ removed before production firmware was built and flashed.
   maximum; cumulative error is 152 us over 25.75 seconds. This records bounded
   BLE scheduling jitter without rate drift and requires timestamp-derived `dt`
   in future fusion work.
+- `ble_mtu_race_4bf00eb.trk` — SHA-256
+  `4a231b54a3c7320fd69cac869b830e94aca8f704f946881b9cdfd21991bfa41f`.
+  This exact-commit stress capture followed six rapid subscribe, first-fragment,
+  and disconnect cycles aimed at the MTU lookup race. Telemetry remained live
+  afterward: the capture contains 1,848 CRC-valid samples with zero packet gaps,
+  resets, send failures, or invalid ACKs, while status records exactly six
+  disconnects and six replays. The test deliberately withheld application ACKs
+  beyond the 10.24-second queue window, so sequences 1,024 through 4,021 were
+  intentionally lost; the single 2,998-sample gap, matching queue-overflow total,
+  demonstrates bounded buffer exhaustion rather than a suspended output task.
 
 `tests/test_trikke_protocol.py` verifies the hashes, parses the captures in
 fragmented chunks, and asserts these signatures so the hardware evidence remains