perf(workspace): share the startup inventory read and defer non-first-card IPC (J1)

Journey J1 / renderer.ipcCallsToFirstCard: 12 -> 7.

- PlaylistsService.getAllPlaylists() shares one in-flight SQLite read
  between concurrent callers (the playlist effect and the XMLTV source
  reconciliation at startup). Settled reads are never reused and every
  SQLite write detaches the pending read.
- getM3uFavoriteChannels() stops re-reading the write-once IndexedDB ->
  SQLite migration receipt once it has been seen.
- StartupDeferralService holds the download list, app update status and
  the dashboard's recent items and favorites until one task after the
  render that reveals the routed content (5 s safety timeout).
  IPTVNATOR_DISABLE_STARTUP_DEFERRAL=1 is the kill switch.
- reconcileEpgSources and the two distinct migration-flag reads stay on
  the critical path: the first is the #1548 revision fence, the second
  are different keys, not duplicates.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

docs/architecture/performance-journeys.md

@@ -179,6 +179,35 @@ harness, which is what the ratchet needs. The main process start
 (`Date.now() - process.uptime()`) is recorded per iteration under
 `evidence.epochs` for cross-checks.
 
+### Startup work before the first card
+
+`renderer.ipcCallsToFirstCard` counts what the renderer asks of the main
+process before the first card. Only work that the first card needs may run
+before it:
+
+- `PlaylistsService.getAllPlaylists()` shares one in-flight SQLite read
+  between concurrent callers (`SharedInFlightRead`): at startup the playlist
+  effect and the XMLTV source reconciliation both read the inventory, and the
+  second caller joins the first read and receives a copy. A settled read is
+  never reused, and every SQLite write detaches the pending read, so a caller
+  that follows a write reads again.
+- `StartupDeferralService` (`@iptvnator/services`) holds work the first
+  screen does not need: the initial download list (the header badge), the
+  app update status (the update card), and the dashboard's recent items and
+  favorites. `app.config.ts` arms it with `STARTUP_WORK_DEFERRAL`, and
+  `AppStartupStatusComponent` opens it one task after the render that
+  reveals the routed content, so released work never lands in the task of
+  the first card. It opens on its own after 5 s if that render never
+  happens. Without the token (tests, other hosts) the gate is open.
+- `IPTVNATOR_DISABLE_STARTUP_DEFERRAL=1` (or `true`) is the short-lived kill
+  switch: the preload exposes it as `startupDeferralDisabled` and the gate
+  starts open. Remove it within two releases of the change that added it.
+- `reconcileEpgSources` stays before the first card on purpose: its
+  completion bumps `EpgSourceSettingsService.revision()`, the fence that
+  keeps XMLTV lookups from returning data of a removed source. Deferring it
+  would let dashboard EPG lookups start first and then be discarded and
+  requeued.
+
 ### Summary schema
 
 ```json

docs/architecture/workspace-dashboard.md

@@ -282,7 +282,11 @@ hooks: `dashboard-hero`, `dashboard-hero-slide` (`data-hero-kind`),
        `updateDate` / `importDate` for sources that have never been used.
 3. `DashboardDataService` is passive on construction. The dashboard feature
    owns the initial reloads for recent items, favorites, and Xtream recently
-   added rows on page entry.
+   added rows on page entry. Recent items and favorites reload through
+   `reloadForPageEntry()`, which on launch waits for `StartupDeferralService`
+   (the first rendered screen, journey J1); their rails keep their scoped
+   skeletons meanwhile, and the Xtream recently-added and TMDB rails follow
+   `globalFavoritesLoaded()` as before. Later page entries find the gate open.
 4. No dashboard-local `Layout` state, no localStorage keys, no migrations.
    Per-rail visibility is the one persisted preference, and it lives in the
    global settings store (`Settings.dashboardRails`), not in a

docs/development/electron-debugging.md

@@ -34,6 +34,7 @@ IPTVNATOR_TRACE_STARTUP=1 pnpm nx serve electron-backend
     - `IPTVNATOR_PERF_COUNT_SQL=1`, together with `IPTVNATOR_PERF_CAPTURE=1`, also counts every SQL statement of the main-process and database-worker connections for `main.sqlStatementsBeforeReadyToShow`; it wraps each statement execution, including every row of a bulk insert, so only the launch journey sets it and the import benchmarks leave it unset
     - `IPTVNATOR_PERF_WORKER_PROFILING=1` enables development/test-only, request-scoped worker receive/work/response-post timestamps, thread CPU, event-loop utilization/delay, count-only playlist serialization/SQLite write/read/deserialization plus Xtream category/content/cache-clear/delete/in-source-search phase events, profiling-only worker cancel-receipt acknowledgements, valid-sample-counted isolate peak memory, and the database worker's idle-only one-shot post-GC heap probe; overlapping database requests are explicitly invalidated instead of misattributed, the performance benchmark sets the flag automatically, and production launches must leave it unset
     - `IPTVNATOR_DISABLE_COMPILE_CACHE=1` disables the main-process V8 compile cache; `IPTVNATOR_COMPILE_CACHE_DIR=<dir>` relocates it. The startup trace reports the outcome as `compile-cache`
+    - `IPTVNATOR_DISABLE_STARTUP_DEFERRAL=1` starts the renderer's startup gate open, so the download list, update status and dashboard recent/favorites reads run before the first screen again (kill switch; see "Startup work before the first card" in `docs/architecture/performance-journeys.md`)
 
 - Settings, portal request/response, and trace payloads must use
   `@iptvnator/shared/logging` or the redacting portal logger before reaching