docs: reconcile ROADMAP candidate pool against shipped examples

Two drift cases found by diffing the pool against examples/gallery.json:

- Removed the 'Gallery coverage follow-ups from the GPv3 review' line:
  every example it names (grease-pencil-rosette, armature-bend,
  text-version-stamp, image-pixels-testcard, vse-cut-list) has shipped,
  and its light-linking subject duplicates the dedicated shipped line
  for light-link-studio.
- Struck through the glTF skinned-mesh export witness entry per the
  pool's convention: it shipped as gltf-skin-roundtrip but was left
  unstruck, so it read as an active candidate.

Every remaining unstruck entry verified genuinely unbuilt (camera DOF,
volumetric scatter, freestyle, chess lathe, dice pair, bicycle wheel,
UV atlas pack, and the two snippet ideas have no matching examples/
directory).

Signed-off-by: fOuttaMyPaint <TMhospitalitystrategies@gmail.com>

Author: TMHSDigital

ROADMAP.md

@@ -97,7 +97,6 @@ Not committed; target list for the next content version. (v0.3.0 shipped the smo
 - Refresh the `slotted-actions-animation` skill against any 5.2 changes
 - Bump `blender_version_min` in the templates if 5.2 APIs are used
 - Additional snippets for asset library scripting, EXR baking, multi-file extensions
-- Gallery coverage follow-ups from the GPv3 review: light-linking (`grease-pencil-rosette`, `armature-bend`, `text-version-stamp`, `image-pixels-testcard`, and `vse-cut-list` shipped first)
 - ~~UV-layer authoring witness~~ **SHIPPED** as `examples/uv-layer-grid/` — `create_grid(..., calc_uvs=True)` silent no-op without a pre-existing UV layer; closed-form UV fill + explicit assignment fallback; dual-panel render (flat texel (0,0) vs neon checker)
 - ~~Image save-format witness~~ **SHIPPED** as `examples/png-exr-alpha/` — float→PNG is RGBA16 and false-unpremultiplies as if associated-alpha (closed-form err 0.98 at RGB 0.02 / a=1/255); OpenEXR preserves float RGBA; byte→PNG is straight RGBA8; `EXR color_mode='RGB'` drops alpha
 - ~~Attribute domain witness~~ **SHIPPED** as `examples/attribute-domain-shear/` — POINT vs CORNER semantics on an 8-wedge pinwheel with one shared hub: CORNER authoring exact (hub corners disagree by face, err ≤ 1e-6), naive per-wedge POINT loop shears to the last write at hub + ring verts, measured deviation == palette closed form (0.751031); byte-identical on 4.5.11 and 5.1.2; crisp-vs-smeared dual pinwheel still — companion to `color-attribute-wheel`
@@ -108,7 +107,7 @@ Not committed; target list for the next content version. (v0.3.0 shipped the smo
 - ~~Vertex weight limit witness~~ **SHIPPED** as `examples/vertex-weight-limit/` — the 4-influence engine cap enforced via the data API (`v.groups` + `VertexGroup.remove` + renormalize); unit sums (measured 3e-8), pose preserved (4.9e-3), exact LBS from the mesh's own deform layer (2.7e-7), Root mount pinned
 - ~~Triangulate + tangent-space witness~~ **SHIPPED** as `examples/triangulate-tangents/` — `calc_tangents` aborts on any ngon (back cap must be an explicit fan); mikktspace matches the edge/UV-delta formula within welding tolerance on smooth fields (2.3e-6 measured); planar UVs on a cylindrical wall collapse tangents onto normals (dot 0.998); `MeshUVLoopLayer` handle dangles across `calc_tangents()` on 4.5 (471 phantom flips, silent exit 0) while the mikktspace math is byte-identical on 4.5.11 and 5.1.2
 - UV-handle lifetime snippet: re-fetch attribute/UV layers by name after any CustomData-reallocating call (`calc_tangents`, `VertexGroup.add`, modifier edits) — held handles dangle silently on 4.5, survive by luck on 5.1 (found authoring `triangulate-tangents`)
-- glTF skinned-mesh export witness: follow-up to `gltf-export-roundtrip` + `vertex-weight-limit` — **SHIPPED** as `examples/gltf-skin-roundtrip/` — `skins[0].joints` names every bone, JOINTS_0/WEIGHTS_0 with unit sums (3e-8), weights bit-exact, rest matrices to 2.4e-07, deformation to 4.8e-07; exporter welds duplicate loops (sorted-multiset comparisons mispair — compare by rest-key); unparented skinned meshes make the exporter bind an armature by name
+- ~~glTF skinned-mesh export witness: follow-up to `gltf-export-roundtrip` + `vertex-weight-limit`~~ **SHIPPED** as `examples/gltf-skin-roundtrip/` — `skins[0].joints` names every bone, JOINTS_0/WEIGHTS_0 with unit sums (3e-8), weights bit-exact, rest matrices to 2.4e-07, deformation to 4.8e-07; exporter welds duplicate loops (sorted-multiset comparisons mispair — compare by rest-key); unparented skinned meshes make the exporter bind an armature by name
 - ~~Degenerate-bevel weld hazard~~ **SHIPPED** as `examples/degenerate-bevel-weld/` — at `offset == min_dim/2` the bevel band collapses to exactly 12 zero-area faces (4 min-axis edges × 3 segments) with 16 coincident verts (4 × (segments+1)) and min_area down 3.2e5×; a stdlib GLB re-parse counts **32 degenerate triangles crossing the export boundary** (MEASURED regression constant) versus 0 for the safe mesh — the loops an engine merge-by-distance welds; byte-identical on 4.5.11 and 5.1.2
 - ~~GAMMA_CROSS blend-curve witness~~ **SHIPPED** as `examples/vse-gamma-cross/` — the cross blends in a gamma-0.5 space: `((1-t)·√A + t·√B)²` with `t = (frame − start)/duration`, never 1 inside the effect; mid-cross dips 0.115 below the sRGB lerp from crimson/teal (closed form (0.341, 0.349, 0.463) confirmed per frame); AgX-default sampling poisons the fit (0.146 red-channel error, `view_transform='Standard'` mandatory); deleting a consumed input orphans-and-deletes the effect — follow-up to `vse-cut-list`
 - Falsy `bpy_prop_collection` trap snippet: an empty collection is falsy, so `editor.strips or editor.sequences` silently falls through to the legacy accessor on an empty timeline — always branch on `hasattr`; likely generalizes across the API (found authoring `vse-cut-list`)