docs(state): audit Open bugs and move resolved rows to Recently closed (#1238)

[lusoris]

Summary

Audited all 19 rows under ## Open bugs in docs/state.md against git log and origin/master to resolve issue #1238.

• Moved four confirmed-fixed bug rows from ## Open bugs to ## Recently closed with merged PR and commit citations:
  • T-UPSTREAM-818-POOLING-ENUM-NO-PERCENTILES-2026-09-03 (fixed by PR feat(core): add percentile temporal pooling to the public C API #1340, commit b43fc4414)
  • T-AI-PTQ-STATIC-QUANT-FORMAT-UNPINNED-2026-09-03 (fixed by PR fix(ai): pin ONNX Runtime static-PTQ output format to QDQ #1306, commit aebb9e19a)
  • T-SVTAV1-HDR-ADAPTER-2026-05-20 (fixed by PR fix(vmaf-tune): address report audit findings #2-#10 and document SVT-AV1-HDR knobs #1296, commit e335857f4)
  • T-VMAFTUNE-PROFILE-REPORT-AUDIT-2026-05-20 (fixed by PR fix(vmaf-tune): address report audit findings #2-#10 and document SVT-AV1-HDR knobs #1296, commit e335857f4)
• Restored two recently closed rows accidentally deleted during dedup in commit 02f9ee455:
  • T-METAL-MOTION-V2-MIRROR-OFF-BY-ONE-2026-09-03 (fixed by PR fix(metal): close out motion_v2 mirror reflect-101 parity and test observability #1294, commit 8d103e3cc)
  • T-UPSTREAM-1564-ADM-CM-GPU-BORDER-AND-ROUNDING-2026-09-03 (fixed by PR fix(feature): row-level rounding and border indexing in integer ADM GPU kernels (ADR-1167) #1224, commit 6c843bb84)
• Deduplicated tombstone comments under ## Open bugs.
• Added changelog fragment changelog.d/changed/docs-1238-state-md-stale-rows.md and regenerated CHANGELOG.md.

Closes #1238.

Type

• feat — new feature
• fix — bug fix
• perf — performance improvement
• refactor — no behavior change
• docs — documentation only
• test — test-only
• build / ci — tooling / infra
• port — cherry-pick from upstream Netflix/vmaf
• sycl / cuda / simd — backend-specific

Checklist

• Commits follow Conventional Commits (the commit-msg hook enforces this).
• make format && make lint is green locally.
• Unit tests pass: meson test -C build.
• If I touched any SIMD/GPU code path, I ran /cross-backend-diff and the worst ULP is ≤ 2.
• If I touched a feature extractor with SIMD/GPU twins, I either updated every twin or listed the gap under "Known follow-ups" below.
• If I added a new .c / .cpp / .cu / .h / .hpp, it has the appropriate license header (see CONTRIBUTING.md).
• If this is a breaking change, the commit message uses ! or BREAKING CHANGE: and the migration path is documented below.
• If this PR adds an ADR, the ADR row lives in docs/adr/_index_fragments/<NNNN-slug>.md and the slug is appended to docs/adr/_index_fragments/_order.txt — do not edit docs/adr/README.md directly (regenerated by scripts/docs/concat-adr-index.sh; see ADR-0221).

Bug-status hygiene (ADR-0165)

• docs/state.md updated in this PR with a row in the appropriate section (Open / Recently closed / Confirmed not-affected / Deferred), OR no state delta: REASON.

Netflix golden-data gate (ADR-0024)

• I did not modify any assertAlmostEqual(...) score in the Netflix golden Python tests.
• If I believe a golden value must change, I have explained why below AND pinged @lusoris for a CODEOWNERS exception.

Deep-dive deliverables (ADR-0108)

• Research digest — no digest needed: documentation-only state ledger audit
• Decision matrix — no alternatives: mechanical ledger sync to master
• AGENTS.md invariant note — no rebase-sensitive invariants: state ledger audit
• Reproducer / smoke-test command — bash scripts/ci/check-state-md-rows.sh
• CHANGELOG fragment — changelog.d/changed/docs-1238-state-md-stale-rows.md
• Rebase note — no rebase impact: documentation and changelog updates only

Reproducer

bash scripts/ci/check-state-md-rows.sh

Known follow-ups

None.

[lusoris]

Rebased onto master (78c9d2b). One conflict, docs/state.md, resolved as a union: master's side added the T-CODE-SCANNING-1243-FIX-2026-09-08 closed row while this branch adds five different bug ids, so both sets are kept and the _Updated: line now records both events. scripts/ci/check-state-md-rows.sh passes: 383 rows, no duplicate ids.

Note that the repo's scripts/dev/resolve-state-md-conflict.py was not used here, deliberately. Its own docstring says it cannot handle a branch that moves rows to Recently closed — it keeps master's copy and you have to redo the move by hand — which is exactly what this PR does. Running it would have silently discarded the point of the change. The moves were verified individually afterwards: three ids now carry a moved to Recently closed tombstone in Open bugs with their canonical row in Recently closed.

One pre-existing inconsistency this PR does not fix and arguably should, since it is in scope for #1238: T-AI-PTQ-STATIC-QUANT-FORMAT-UNPINNED-2026-09-03 has a canonical row in Open bugs at docs/state.md:359 and a tombstone at :361 saying it moved to Recently closed. That contradiction is already on master (same two lines), and this branch only rewords the tombstone rather than completing the move.

[lusoris]

Heads-up: this PR's work is in the pre-rc.1 fixing train, #1425 — its commits were replayed onto the train individually, with each conflict resolved by hand and recorded in the commit messages.

Leaving it open rather than closing it, because unlike its siblings this branch adds no ADR file, so it is not tripping the ADR Collision Guard on #1425 and nothing is blocked by it. Close it whenever you like, or leave it as a cross-check until #1425 merges.

[lusoris]

Absorbed in effect by the rc.1 train (#1425). The commit moves four rows out of Open bugs; on the train those rows are already in ## Recently closed and already marked closed, so cherry-picking it resolves to a no-op against the train (the only conflicting hunks are deletions of rows the train has already relocated). check-state-md-rows.sh reports no duplicates. Closing — reopen if you disagree.