docs: record fail-closed accuracy challenger

backend/README.md

@@ -519,6 +519,13 @@ false-negative counts, but has lower precision than the previous balanced
 model; operators must review and persist QA/QC rather than treating detections
 as ground truth.
 
+The latest coverage-aligned rerun of this exact profile measured mean precision
+`0.6141`, recall `0.6062` and F1 `0.6069` over Mol Achterbos, Donk, Gompel and
+Postel plus Retie, Turnhout and Westerlo. The three pure-empty controls remained
+at zero detections. A reviewed six-AOI fine-tuning challenger reached mean F1
+`0.6248` but remained inactive because it produced two false detections in the
+Postel-bos empty control.
+
 For model-quality calibration, run the confidence sweep wrapper:
 
 ```bash

backend/tests/test_sprint155_detection_operator_profiles.py

@@ -20,12 +20,14 @@ def test_detection_operator_profiles_define_explicit_yolo_candidates_and_promote
     assert "defaultApproved: true" in source
     assert "promotionRecommendation: 'promote_candidate'" in source
     assert "positiveSampleCount: 7" in source
-    assert "f1: 0.5824578631584316" in source
+    assert "precision: 0.6140895327792112" in source
+    assert "recall: 0.6062221049337548" in source
+    assert "f1: 0.6068607646002744" in source
     assert "f1: 0.5432865390636915" in source
     assert "maxBackgroundDetections: 0" in source
     assert "lege-achtergrondtest is geslaagd" in source
-    assert "1.571 minder gemiste gebouwen" in source
-    assert "onterecht gevonden objecten" in source
+    assert "Postel blijft met 47,5% F1" in source
+    assert "controlekandidaat en niet als grondwaarheid" in source
 
 
 def test_detection_lab_surfaces_profiles_as_deliberate_operator_actions() -> None:

backend/tests/test_sprint193_end_user_workbench.py

@@ -67,7 +67,8 @@ def test_visible_ai_and_quality_labels_are_end_user_facing() -> None:
     providers = read("frontend/src/components/providers/ProviderPanel.tsx")
 
     assert "Aanbevolen controleprofiel kleine gebouwen" in profiles
-    assert "Controleer wel extra op onterecht gevonden objecten" in profiles
+    assert "Postel blijft met 47,5% F1" in profiles
+    assert "controlekandidaat en niet als grondwaarheid" in profiles
     assert "qualityStatusLabel" in quality
     assert "nog niet uitgevoerd" in quality
     assert "Nog geen bestand gekozen." in export_preview