perf(inventory): lazy list, photo thumbnails, indexes, debounced reload

Scale the local inventory to 10k+ varieties:
- Render the list with ListView.builder over a flattened header/item model
  instead of building every tile upfront.
- Store a small regenerable JPEG thumbnail per photo (schema v14) and use it
  for the 48px list avatar; full bytes stay for offer image hosting. Existing
  photos are backfilled lazily at startup. Thumbnail is local-only (excluded
  from CRDT sync and backups by the JSON codec).
- Add indexes on varieties(is_deleted,is_draft), attachments(parent_type,
  parent_id,kind), lots(variety_id) via @TableIndex.
- Debounce watchInventoryView (~250ms) so a burst of table writes triggers one
  reload, not seven.
- cacheWidth/cacheHeight on the list avatar decode.
- Scale test raised 3k -> 10k; migration test v13 -> v14.

apps/app_seeds/test/data/inventory_scale_test.dart

@@ -19,8 +19,8 @@ void main() {
   });
   tearDown(() => db.close());
 
-  test('the inventory view loads 3000 varieties quickly', () async {
-    const count = 3000;
+  test('the inventory view loads 10000 varieties quickly', () async {
+    const count = 10000;
     for (var i = 0; i < count; i++) {
       final id = await repo.addQuickVariety(
         label: 'Variety $i',
@@ -43,11 +43,12 @@ void main() {
     sw.stop();
 
     expect(view.items, hasLength(count));
-    // Generous ceiling for CI; typical local runs are well under 500ms. The
-    // point is to catch an accidental N+1 regression, not to micro-benchmark.
+    // Generous ceiling for CI (the point is to catch an N+1 regression, not to
+    // micro-benchmark). Includes the ~250ms debounce on the first emit; typical
+    // local runs are well under 2s for 10k rows with the v14 indexes.
     expect(
       sw.elapsedMilliseconds,
-      lessThan(3000),
+      lessThan(6000),
       reason: 'inventory view took ${sw.elapsedMilliseconds}ms for $count rows',
     );
   });