diff --git a/android/play-store/closed-test.md b/android/play-store/closed-test.md
index 10251d2..3a34c31 100644
--- a/android/play-store/closed-test.md
+++ b/android/play-store/closed-test.md
@@ -5,7 +5,7 @@ least 12 testers to remain opted in to the closed test for 14 continuous days
 before production access can be requested.
 
 This requirement was rechecked against the official Play Console Help article
-on 2026-08-25. The Console remains the source of truth for the account's actual
+on 2026-08-26. The Console remains the source of truth for the account's actual
 eligibility and the date on which production access can be requested.
 
 ## Before inviting testers
diff --git a/android/play-store/release-checklist.md b/android/play-store/release-checklist.md
index 9511759..dde077e 100644
--- a/android/play-store/release-checklist.md
+++ b/android/play-store/release-checklist.md
@@ -213,9 +213,9 @@ block publishing changes, including Store Listing, Pricing, and Distribution.
 On 2026-08-25 the existing **Closed testing - Alpha** track was inactive with
 zero of four setup tasks complete. Countries and testers were not selected, and
 the track had no release. The current official requirement for this personal
-developer account was rechecked as at least 12 testers continuously opted in
-for at least 14 days before applying for Production access. No Closed or
-Production release was created during this inspection.
+developer account was rechecked again on 2026-08-26 as at least 12 testers
+continuously opted in for at least 14 days before applying for Production
+access. No Closed or Production release was created during this inspection.
 
 ## Publishing
 
diff --git a/docs/performance.md b/docs/performance.md
index 047bca0..b7971b7 100644
--- a/docs/performance.md
+++ b/docs/performance.md
@@ -461,6 +461,17 @@ host. It is not a production capacity claim and does not justify a hard startup
 guard. The much slower hot-scope result is expected contention on one database
 row; ordinary email/IP policy traffic distributes attempts across many scopes.
 
+The same checked-in benchmark was repeated on 2026-08-13 with the same explicit
+one-CPU, PostgreSQL 18.4, 5,000-call and concurrency-20 boundaries. The hot
+scope completed in 12,625.396 ms (396.03 operations/s); 5,000 distributed
+scopes completed in 852.909 ms (5,862.29 operations/s). Both profiles reported
+5,000 successful calls, zero failures, zero unexpected limited calls, exactly
+5,000 persisted increments, and zero run-owned buckets after cleanup. The
+generated JSON is retained under the ignored local directory
+`output/rate-limit/20260813073442-3329179/`. This repetition strengthens the
+implementation and cleanup evidence; it still does not select or approve the
+product-facing thresholds.
+
 ## Observed local comparison
 
 Observed on 2026-07-18 with Docker Engine 29.6.2 on the recorded 32-CPU,