fix: stage the two newest heroes as designed objects

[TMHSDigital]

Summary

Second art-direction pass on the two newest examples. The first re-stage (#93) fixed the measured defects, but the cards still read flat next to siblings that wrap their witness in a designed object (rockets, mech arm, monitor prop, angled panels). This pass dresses both staged scenes as physical objects. Check logic untouched; render-path and docs only.

vse-gamma-cross — the floating chart becomes a calibration lightbox: the six exact data swatches are backlit chips seated in bezels on a bolted, cap-railed steel swatch board with rear struts, shot at a 3/4 angle with cast shadows. Swatch faces stay pure emission at the exact closed-form values; the authentic-vs-naive adjacency pair (impostor in hazard orange) remains the centerpiece — a naive-lerp cross would make the pair identical, so the still still visibly breaks with the contract.

light-link-studio — the bare spheres become museum specimens: each on a turned, beveled pedestal with a physical LINKED/UNLINKED placard, and a warm emissive floor inlay under the hero only — the linked light's footprint, with the decoy's disc staying dark metal.

Notable finding during the pass

The re-stage exposed a real coupling in light-link-studio: a flatter camera angle samples a brighter, more fill-facing point on the decoy and eroded the linked ratio to 3.1x (margin floor 3.2x). That draft was reverted; the coupling is documented in a build_studio comment. The pedestal now shades floor bounce onto the decoy, improving the 4.5 ratio (3.4x → 3.7x).

Synced measured values (scene changed, values drifted)

Metric	5.1.2	4.5.11	Was (5.1 / 4.5)
linked ratio (gate ≥3.0x)	3.6x	3.7x	3.6x / 3.4x
unlink rise (gate ≥50%)	233%	238%	206% / 213%
hero drift (gate ≤5%)	0.3%	0.1%	0.0% / 0.0%
EEVEE probe	3.8x	3.8x	3.7x

Updated consistently in the example READMEs, gallery.json, and regenerated pages. vse-gamma-cross check values unchanged (samples=9 q_err=2.93e-03 lerp_mid_dev=0.115).

Contact-sheet gate (measured)

• docs/gallery/contact-sheets/vse-gamma-cross-contact-sheet.webp
• docs/gallery/contact-sheets/light-link-studio-contact-sheet.webp

Criterion	vse-gamma-cross	light-link-studio	Calibration band
Mean luminance	115.2	127.9	100.1–143.4
Wall pool R−B	+20.8	+26.2	+8…+58
Subject fill	83% vertical	82% width	70–90%
Thumbnail legibility	labels + pair read at 400 px	placards read at 400 px	—

Neither candidate sorts as the odd one out.

Test plan and evidence labels

Proven by live run (local headless binaries):

• .scratch/blender-5.1.2-windows-x64/blender.exe (self-reports Blender 5.1.2, hash ec6e62d40fa9) and .scratch/blender-4.5.11-windows-x64/blender.exe (self-reports Blender 4.5.11 LTS, hash 4db51e9d1e1e)
• Both examples check-only, exit 0 on both versions, with the printed values in the table above (parent re-ran all four check runs on the shipped tree, not only during authoring)
• light-link EEVEE luminance probe re-run on the new scene: 3.8x on both binaries
• Render path also exercised on 4.5 (full --output still), visually identical to 5.1

Established by inspection only:

• Generated docs/gallery/index.html cards and both detail pages read after regeneration: witness bullets render as <li>, no truncation or duplicated words
• blender-smoke.yml steps unchanged

CI: the two blender-smoke legs (4.5 LTS, 5.1) on this PR remain the authoritative cross-version evidence.