What

Collapse the three reset-reentry tier-rollup summary renderers —
..._rebuild_summary, ..._rebuild_reentry_restore_summary,
..._rebuild_reentry_restore_rerererestore_summary — onto one
_tier_summary_base + a _TierSummarySpec NamedTuple + three per-tier specs.
Each public function is now a thin wrapper with its exact original signature.

Why

The three were structural clones (AST-identical after literal-blanking),
differing only in:

• the status key they read,
• the per-tier status tokens, and
• the reworded prose.

The status-string render skeleton was triplicated — the same drift risk that
already produced two real bugs earlier in this campaign (the phase-5 floor and
freshness-flag fixes). A future status branch now lands once, not three times.

Proof of byte-identical behavior

1. Exhaustive branch differential — 168 cases per tier (504 total) spanning
   every status branch, both hotspot fallbacks, blocked-reason present/absent,
   and label / hotspot-label combinations: real(input) == base(input, spec)
   for every case.
2. Composer golden contract — zero diff (tests/test_composer_golden_contract.py,
   47 passed).

The prose templates ({label} / {hotspot_label} str.format strings) were
extracted from the source AST, never hand-typed, so the contract text is exact.

Verification

• uv run pytest -q → 2539 passed, 2 skipped
• uv run mypy src/operator_trend_closure_forecast_reset_controls.py → clean
• uv run ruff check src/ tests/ → clean

Net 136 insertions / 138 deletions (near line-neutral by design — these
renderers are mostly prose; the win is eliminating the triplicated render
skeleton and its drift risk, consistent with the phase-6 collapses).