fix(replicas): never REFRESH with copy_data=true after a data copy

A second logical copy unique-conflicts on existing PKs and leaves
tablesync workers stuck, so live UPDATEs never apply on slaves.

Co-authored-by: Cursor <cursoragent@cursor.com>

deploy/sync-replica-data.sh

@@ -322,7 +322,7 @@ repair_replica() {
     run_psql "$(libpq_dsn "$replica_dsn")" -v ON_ERROR_STOP=1 -q \
       -c "ALTER SUBSCRIPTION \"${sub}\" ENABLE"
     run_psql "$(libpq_dsn "$replica_dsn")" -v ON_ERROR_STOP=1 -q \
-      -c "ALTER SUBSCRIPTION \"${sub}\" REFRESH PUBLICATION" || true
+      -c "ALTER SUBSCRIPTION \"${sub}\" REFRESH PUBLICATION WITH (copy_data = false)" || true
   done
 
   echo "  repair done"