MuJoCo Web Viewer: Implement parallel chunked model downloading and in-place model reloading.

This change introduces parallel chunked downloading for large model files (.mjb) directly into WASM linear memory, enables model loading without full page reloads and reduces memory overhead.

Key changes:
- Implement a chunked model endpoint in the Python web server to support range requests.
- Add parallel chunked fetching in the frontend with retry logic and a single-fetch fallback.
- Increase initial WASM memory to 3 GB to accommodate large models and prevent heap fragmentation.
- Support in-place model reloading in the C++ client, including texture cache invalidation when the Filament context is recreated.
- Display a model download progress bar and model parsing/loading banner to the UI.
- Fixes model drag and drop (caused by typo in sessionId, corrected to session_id).

PiperOrigin-RevId: 960249236
Change-Id: Icac89e6a4ca099882aaf9b111744c1b6ab0cc6c1

python/mujoco/experimental/studio/web/state_payload.h

@@ -20,7 +20,7 @@
 //
 // The arguments come from the Python process:
 //
-//   * model       : fetched once over HTTP as /model.mjb; its runtime-mutable
+//   * model       : fetched once over HTTP as /model; its runtime-mutable
 //                   parts (opt/vis/stat) are re-sent in the render state block.
 //   * data        : streamed as the physics state vector (mjSTATE_INTEGRATION);
 //                   the browser recomputes the rest via mj_setState/mj_forward.
@@ -64,7 +64,7 @@ struct StatePayloadHeader {
   uint16_t version = kStatePayloadVersion;
   uint16_t nblocks = 0;
   // CRC32 of the model's MJB bytes. When this changes, the browser must
-  // refetch /model.mjb before applying any further state.
+  // refetch /model before applying any further state.
   uint32_t model_crc32 = 0;
 };
 static_assert(sizeof(StatePayloadHeader) == 12);