streaming: near LODs first (walkable ring, LOD0, LOD1, LOD2, then far); exact passage trims (L78)

- Headless and Unreal schedulers drain the collision-critical ring, then LOD0-2, then far LODs
  at low priority. LOD0-2 ready (W8 medians): SW 4.36->1.76, FI 6.89->5.26, TN 4.88->3.22,
  VS 2.71->1.56 s. Whole view roughly unchanged (FI/VS slightly later).
  KNOWN RISK: far tiles wait until every near tile is ready, so far LODs may starve while
  moving fast (to address next).
- Exact composer trims: tile-level upper-mouth candidate list, segment-index min/max rejection
  before lower_bound, 64-entry thread-local value-noise corner cache. Passage share 65 -> 63%.
- World exact: 24/24 hash matrix + pinned L75 owner hash. Fast tier PASS. Linux UE host builds.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016dhxM43dnWvUu2B5Ph1iG1

Standalone/Source/StreamBenchmarkPipeline.h

@@ -5061,6 +5061,7 @@ namespace StandaloneStreamBenchmark
                 *VoxelForgeCore::Linux77Profiler::RouteCount*2>,9> L77DispatchByLOD{};
         std::array<std::uint64_t,9> L77GridInteriorByLOD{},L77GridHaloByLOD{};
         std::array<double,9> L77LODReadySeconds{};
+        std::array<double,9> LODReadySeconds{};
         std::array<std::uint64_t,9> SkirtTrianglesByLOD{};
         std::array<std::uint64_t,9> TileCountByLOD{};
         std::array<std::uint64_t,9> ClassifiedAirTilesByLOD{},ClassifiedSolidTilesByLOD{};
@@ -5129,6 +5130,9 @@ namespace StandaloneStreamBenchmark
                     const std::size_t LOD=static_cast<std::size_t>(Tile.Key.Level);
                     const auto& TileDiagnostics=Tile.CoreDiagnostics;
                     ++TileCountByLOD[LOD];
+                    if(TileCountByLOD[LOD]==ExpectedTilesByLOD[LOD])
+                        LODReadySeconds[LOD]=std::chrono::duration<double>(
+                            Tile.CompletedAt-ColdStart).count();
                     if(bLinux77Profile)
                     {
                         const auto& D=TileDiagnostics;
@@ -5551,6 +5555,23 @@ namespace StandaloneStreamBenchmark
                 }
             }
         }
+        double NearLODReadySeconds=0.0;
+        uint64 NearLODExpected=0,NearLODCompleted=0;
+        for(std::size_t LOD=0;LOD<=2;++LOD)
+        {
+            NearLODExpected+=ExpectedTilesByLOD[LOD];
+            NearLODCompleted+=TileCountByLOD[LOD];
+            if(ExpectedTilesByLOD[LOD]!=0)
+                NearLODReadySeconds=std::max(NearLODReadySeconds,LODReadySeconds[LOD]);
+        }
+        std::printf("[WorldAssemblerNearLODReady] lods=0-2 tiles=%llu/%llu cold_ready_s=%.6f\n",
+            static_cast<unsigned long long>(NearLODCompleted),
+            static_cast<unsigned long long>(NearLODExpected),NearLODReadySeconds);
+        for(std::size_t LOD=0;LOD<TileCountByLOD.size();++LOD)
+            if(ExpectedTilesByLOD[LOD]!=0)
+                std::printf("[WorldAssemblerLODReady] lod=%zu tiles=%llu cold_ready_s=%.6f\n",
+                    LOD,static_cast<unsigned long long>(TileCountByLOD[LOD]),
+                    LODReadySeconds[LOD]);
         Streamer.Stop();
         Streamer.DumpColdTimelineTrace();
         bool bScreenSpaceGatePassing=true;