feat: GPU optimization — lighting, viewport CA, benchmark mode, 531 FPS

- Resolution: 8x8 world cells, 2x2 UI cells (UI_SCALE=4)
- GPU lighting: vertex-shader computed, light source list buffer (max 64)
  instead of O(N×R²) grid scan, O(N×S) per cell
- Viewport-aware CA: iterate only active chunks, not all 250×250
- Flat array entity/item/shadow maps instead of HashMaps
- Flat 128-entry ASCII atlas array instead of HashMap lookup
- Partial grid upload: viewport + 30-cell margin only
- Pre-allocated viewport arrays in renderer structs (zero alloc/frame)
- Skip CPU lighting for GPU modes (pass None)
- Benchmark mode: --mode benchmark with per-subsystem timing
- GpuLightSource struct, light_count in push constants
- gather_sources_in_range() for viewport-scoped source gathering

Benchmark (600 ticks, release):
  Graphics: 531 FPS (was 386, +38%), render 1013us (was 1699us, -40%)
  ASCII:    402 FPS (was 313, +28%), render 1502us (was 2346us, -36%)

All 171 tests + 14 scenarios pass.

tests/slime.rs

@@ -28,7 +28,7 @@ fn slime_spawns_correctly() {
     assert!(slime.is_some(), "slime should exist after spawn");
     let sl = slime.unwrap();
     assert!(sl.alive, "slime should be alive");
-    assert_eq!(sl.max_health, 25.0, "slime max health should be 25");
+    assert_eq!(sl.max_health, 65.0, "slime max health should be 65");
 }
 
 #[test]
@@ -46,11 +46,11 @@ fn slime_takes_damage_and_dies() {
         .find(|e| e.kind == "Slime")
         .unwrap()
         .id;
-    s.perform_action(&AiAction::DamageEntity { id, amount: 25.0 });
+    s.perform_action(&AiAction::DamageEntity { id, amount: 65.0 });
     s.step(1);
     let entities = s.get_entities();
     let sl = entities.into_iter().find(|e| e.id == id).unwrap();
-    assert!(!sl.alive, "slime should die after 25 damage");
+    assert!(!sl.alive, "slime should die after 65 damage");
     assert_eq!(sl.kind, "Corpse", "dead slime should become corpse");
 }