diff --git a/workplans/CUST-WP-0074-qualified-wait-states.md b/workplans/CUST-WP-0074-qualified-wait-states.md
index 27063de..e237123 100644
--- a/workplans/CUST-WP-0074-qualified-wait-states.md
+++ b/workplans/CUST-WP-0074-qualified-wait-states.md
@@ -253,7 +253,7 @@ edges appear once T02's indexing is deployed on the primary.
 
 ```task
 id: CUST-WP-0074-T05
-status: progress
+status: done
 priority: high
 needs_human: false
 depends_on: [CUST-WP-0074-T01, CUST-WP-0074-T02]
@@ -322,6 +322,17 @@ third form; do not force it into `needs_human`.
 Long tail beyond the gate roots (about 30 workplans) is left to the C-39
 sweep: `make check-consistency --all` gives the queue.
 
+**Long tail closed 2026-09-29** — a direct scan of every repo's workplan files
+found 248 wait tasks, 223 qualified (`depends_on`, `needs_human`,
+`blocked_on` or `wait_until`). The 25 unqualified are all out of scope: tasks in
+`flavor: residual` workplans (APPROVAL-WP-0002, HUB-WP-0011, TEN-WP-0012,
+RAIL-FAB-WP-0028/0031, WARDEN-WP-0039, WHITEHAT-WP-0006/0008, REUSE-WP-0023),
+two stale waits in finished sand-boxer plans (SAND-WP-0003-T09,
+SAND-WP-0005-T06; owner cleanup), and two `PREFIX-WP-NNNN` template
+placeholders. Zero unqualified waits remain in open non-residual workplans.
+`consistency_check.py --all` exceeded 30 minutes with no output, so it was
+not used; the scan replaced it.
+
 ## Correct the mis-blocked workplans
 
 ```task