Slate v2 Doc Stack Consolidation

Goal

Tighten the live docs/slate-v2 stack so active queue docs stop mixing with legacy/reference material.

The zero-regression diff seed in 2026-04-11-slate-v2-ime-mobile-browser-file-ledger.md is EMPTY, so there are no hidden unreviewed browser/input files left in the current origin/main...HEAD diff scope.
Every row in the behavior/parity ledger is reviewed.
The remaining open parity work is external/tooling-limited:
- Android keyboard features
- broader iOS Safari composition/focus
slate-batch-engine.md and chunking-review.md are useful history, not live queue owners.

demoted batch-engine and chunking docs to explicit historical/reference status
merged the live normalization read into normalization-reference.md
archived:
- api-drift-perf-register.md
- api-drift-perf-scoreboard.md
- api-surface-keep-cut-register.md
- normalization-baseline-matrix.md
- normalization-phase-boundary-note.md
- normalization-reopen-verdict.md
patched live and historical references so the moved docs still resolve
made the front-door docs point at the live parity, verdict, and normalization owners instead of implying batch-engine/chunking are still active candidates
moved legacy/reference docs out of the live root:
- architecture-contract.md -> reference/
- live-shape-register.md -> reference/
- normalization-reference.md -> reference/
- chunking-review.md -> archive/
- slate-batch-engine.md -> archive/
- oracle-harvest-ledger.md -> archive/
- full-replacement-blockers.md -> archive/
- perf-gate-package.md -> archive/
merged blocker ownership into release-readiness-decision.md
merged perf-gate ownership into replacement-gates-scoreboard.md
live docs/slate-v2/*.md root is now down to 9 owner docs