skip: passage proof bounds include wall/floor relief (L75 soundness fix); FI -7.6%

L74's passage bound omitted the 0.21 m wall relief and 0.20 m floor relief, so its whole-tile
proof was unsound in principle (no tested view hit it). Passage support is now bounded per route
segment by four parameter-subinterval boxes using Catmull-Rom derivative extrema, local max
radius and relief padding. Stratum slicing only at LOD0-2 in FI/VS. Legacy sub-tile block
certificates are audit-only (runtime version was slower). TN regression (-80,-20,-3218) kept.

Exact full-grid vs skip hashes on every tile, 4 views, Clang/GCC, W1/W8/W12; control caught;
80-camera sweep 0 violations. Whole view vs c67b200: SW -1.3%, FI -7.6%, TN -2.3%, VS +0.2%.
Compiles in the Linux UE host (VFHostEditor, 27/27 actions).

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

Standalone/Tools/verify/verify.py

@@ -125,10 +125,11 @@ def owner_hash_baseline(recipe: str, style: int) -> tuple[Path, str]:
         return (BASELINES / "owner-hash/canonical/owner-hash-surface-world-style1-v212.baseline",
                 "L67-v212")
     if recipe == "floating-islands" and style == 2:
-        # L74 disables a legacy sub-tile shortcut after its falsifier found false solids.
-        # Keep the old fixture intact and pin the restored full-grid density signature.
-        return (BASELINES / "owner-hash/canonical/owner-hash-floating-islands-style2-l74-fullgrid.baseline",
-                "L74-full-grid")
+        # L75 explicitly materializes every reference density sample. The older L74
+        # signature omitted samples skipped inside native uniform blocks; retain that
+        # fixture and pin the current full-grid signature separately.
+        return (BASELINES / "owner-hash/canonical/owner-hash-floating-islands-style2-l75-fullgrid.baseline",
+                "L75-full-grid")
     raise VerificationFailure(f"no pinned owner hash baseline for {recipe} style {style}")