diff --git a/tests/domain/sap10_calculator/test_real_cert_sap_accuracy.py b/tests/domain/sap10_calculator/test_real_cert_sap_accuracy.py
index 24dde9482..0bc3a95a7 100644
--- a/tests/domain/sap10_calculator/test_real_cert_sap_accuracy.py
+++ b/tests/domain/sap10_calculator/test_real_cert_sap_accuracy.py
@@ -907,12 +907,34 @@ _EXPECTATIONS: Final[tuple[RealCertExpectation, ...]] = (
     # (Elmhurst Jan (83)=56.9 W vs our real-orientation 123.4 W). Verified:
     # forcing our engine to ALSO use a single synthetic North window (i.e.
     # replicating the build's simplification, not a cert or engine fact)
-    # plus the forced Table 11 above reproduces 72.62 -- matches Elmhurst's
-    # 72 almost exactly. Conclusion: the calculator is faithful (it already
-    # matches lodged exactly using the cert's REAL window orientations);
-    # both residuals are Elmhurst-worksheet-reconciliation artifacts (a
-    # deliberate spec choice + a build-entry simplification), not mapping
-    # or calculator bugs. PINNED to the observed exact match vs lodged.
+    # plus the forced Table 11 above reproduces 72.62 (precise Elmhurst
+    # value is 72.0681, not the rounded "72" -- true residual after Table
+    # 11 + windows was ~0.55, not ~1.6). Third cause found by walking the
+    # rest of the worksheet monthly: (43) Average daily hot water use is
+    # 117.4056 L/day (Elmhurst) vs our 110.7281 L/day -- SAME occupancy
+    # (42)=2.4234 both sides, so the gap is in shower/bath demand, not
+    # occupancy. Root cause: cert's `mixer_shower_count` is unlodged (raw
+    # `shower_outlets=None`), which `_mixer_shower_flow_rates_from_cert`
+    # maps to 0 showers (see its docstring: gov-API "no shower_outlets
+    # entry" convention, validated on cert 2225) -- but Elmhurst's build
+    # shows a "Non-electric shower" fixture (1 vented mixer @ Table J4),
+    # since its UI silently defaults one when the build script never
+    # explicitly configured "0 showers" to match the cert. Verified: a
+    # single 7 L/min vented mixer reproduces Elmhurst's (42a) monthly
+    # series EXACTLY (4 d.p., all 12 months) and closes hot water to
+    # 2616.04 vs Elmhurst's 2616.03 (Δ0.005). Corpus-wide check (170
+    # certs sharing "bath present, no shower data lodged") shows NO
+    # systematic bias (mean dSAP +0.12, median -0.04, 46.5% over-rated) --
+    # confirms the mapper's 0-showers convention is correct on average;
+    # this is a THIRD build-entry artifact, not a mapper bug. Combining
+    # all three (windows + Table 11 + 1 shower): SAP 72.4567 vs Elmhurst's
+    # precise 72.0681 -- residual 0.39, ~91% of the original 4.26-point
+    # gap (76.33 vs 72.07) now precisely explained. Conclusion: the
+    # calculator is faithful (matches lodged 76 exactly using the cert's
+    # REAL windows and REAL shower data); all three residuals are Elmhurst
+    # -reconciliation-build artifacts or deliberate spec choices, not
+    # mapping or calculator bugs. PINNED to the observed exact match vs
+    # lodged.
     RealCertExpectation(
         schema="RdSAP-Schema-21.0.1",
         sample="uprn_100061275133",