docs(E3,E4): pro aliquibus locis, not sanctoral gaps

Both entries reasoned correctly that these feasts are not in the universal
calendar, then recommended adding them to CLAUDE.md's list of sanctoral
entries MISSING from colitur. Those cannot both be true, and acting on the
recommendation would add a particular-calendar feast to the universal core
-- what binding decision 2 exists to forbid.

The Missal classifies both itself, under "Missae pro aliquibus locis":
scan1.txt:52019 for the Miraculous Medal, 48793 for St John before the
Latin Gate. Masses for some places.

So they are not gaps. They are particular feasts colitur correctly excludes
and an overlay would supply, the shape poland.ini and benedictine.ini
already demonstrate; the ordos carry them because those ordos serve
communities that keep them. Verdicts unchanged, now resting on the Missal's
own words rather than on background knowledge of the apparition.

1 files changed, 32 insertions, 0 deletions

diff --git a/data/ef/expected-divergences-efdotorg.sexp b/data/ef/expected-divergences-efdotorg.sexp
index 7c9eea7..addecfe 100644
--- a/data/ef/expected-divergences-efdotorg.sexp
+++ b/data/ef/expected-divergences-efdotorg.sexp
@@ -71,6 +71,38 @@
     (note "Our Lady of Ransom (24 September) is Commemoration_only-status data; the September Ember Saturday always wins the day outright. Fires once, 2026-09-24 (2025 and 2027's own 24 September already match colitur without needing this entry -- the Ember day is listed first there).")
     (expected_rows 1))
 
+
+  ; CORRECTED 2026-08-24 (coordinator): this entry reasoned correctly and
+  ; then recommended the opposite. It concludes, rightly, that the feast was
+  ; "never added to the GENERAL calendar" -- and then proposes it as a
+  ; candidate for CLAUDE.md's list of sanctoral entries MISSING from
+  ; colitur's universal data. Those cannot both be true, and acting on the
+  ; recommendation would be a REGRESSION: adding a particular-calendar feast
+  ; to the universal core is precisely what binding decision 2 forbids
+  ; ("any community's proper is an overlay, never core").
+  ;
+  ; The Missal settles it from the primary source rather than from general
+  ; knowledge of the 1830 apparition. BOTH feasts named in E3 and E4 appear
+  ; in the Missal under its own heading "Missae pro aliquibus locis" --
+  ; Masses FOR SOME PLACES:
+  ;     docs/research/scan1.txt:52019, 52077, 52129
+  ;        "Missae pro aliquibus locis (B. M. V. Immaculatae a sacro Numismate)"
+  ;     docs/research/scan1.txt:48793
+  ;        "Missae pro aliquibus locis (S. Ioannis Ap. et Ev. ante Portam Latinam)"
+  ; This is the Missal's OWN classification, and it is the same structural
+  ; test the register already relies on for universality: the calendarium
+  ; carries no "pro aliquibus locis" marker in its own lines, while the
+  ; wider Missal carries the marker elsewhere for exactly this class of
+  ; feast.
+  ;
+  ; So E3 and E4 are NOT sanctoral gaps and are NOT candidates for any
+  ; missing-entry list. They are particular-calendar feasts that colitur
+  ; correctly excludes from the universal core, and that belong in an
+  ; overlay -- the shape data/ef/examples/poland.ini and benedictine.ini
+  ; already demonstrate. The ordos carry them because those ordos serve
+  ; communities that keep them. Verdict [colitur] is unchanged and now rests
+  ; on the Missal's own words.
+
   ; E3 -- colitur correct (a genuine sanctoral DATA GAP, not a rule bug).
   ; "Our Lady of the Miraculous Medal" (27 November) does not appear
   ; anywhere in data/ef/sanctoral.sexp (`grep -i medal` / `grep -i