docs(todo): waive the TWFE full-dummy dedup row (premise resolved by #655)

[igerber]

Summary

• Move the TWFE HC2/HC2-BM full-dummy dedup row from Actionable Backlog to Deferred -> Won't-fix/waived (decisions on the record), recording the scoping outcome.
• The row's shared-helper half already shipped in refactor(utils): shared FE-dummy design build across DiD/MPD/TWFE #655: both the TWFE full-dummy branch and DiD's fixed_effects= branch delegate dummy construction, the drop-first convention, FE column naming, and the duplicate-term backstop to the single build_fe_dummy_blocks + validate_design_term_names implementation in utils.py - the FE-naming / survey-behavior drift risk the row targeted is gone.
• The remaining full-delegation option (routing TWFE.fit through DiD's branch with cluster-default threading) would touch TWFE's user-visible result surface (coefficient-dict keys, cluster-label conventions, warning text) to dedup ~4 genuinely estimator-specific assembly lines; waived on cost/benefit.

Methodology references (required if estimator / math changes)

• Method name(s): N/A - no methodology changes (TODO.md bookkeeping only)
• Paper / source link(s): N/A
• Any intentional deviations from the source (and why): None

Validation

• Tests added/updated: No test changes (docs-only row move)
• Backtest / simulation / notebook evidence (if applicable): N/A

Security / privacy

• Confirm no secrets/PII in this PR: Yes

---

Generated with Claude Code

🤖 Generated with Claude Code

https://claude.ai/code/session_01Lbd6nqWmg4snvvBmegwqiw

[github-actions]

Overall Assessment

✅ Looks good — no unmitigated P0/P1 findings.

Executive Summary

• PR changes only TODO.md, moving the TWFE HC2/HC2-BM full-dummy dedup item from Actionable Backlog to documented waived work.
• No estimator, variance, weighting, inference, defaults, or identification code changes are introduced.
• The waiver is consistent with current source: TWFE and DiD both use build_fe_dummy_blocks() and the final duplicate-name guard.
• The Methodology Registry already documents TWFE’s inline full-dummy route and why it is distinct from DiD’s fixed_effects= surface.
• No security, performance, or test-blocking issues found.

Methodology

Finding: None.

• Severity: None
• Impact: No methodology behavior changes. The referenced methods are TwoWayFixedEffects HC2/HC2-BM and DifferenceInDifferences(fixed_effects=...), but the PR only changes TODO bookkeeping.
• Concrete fix: None required.

Cross-check: Registry documents DiD full-dummy routing and TWFE’s inline full-dummy route, including TWFE-specific cluster/result-surface behavior at docs/methodology/REGISTRY.md:L3427-L3428. Current code shares dummy construction and name validation via diff_diff/utils.py:L220-L265, diff_diff/twfe.py:L360-L381, and diff_diff/estimators.py:L558-L570.

Code Quality

Finding: None.

• Severity: None
• Impact: No executable code changed; no new inline inference, NaN guard, parameter propagation, or control-group logic risk.
• Concrete fix: None required.

Performance

Finding: None.

• Severity: None
• Impact: No runtime path changed. The waived full-delegation option is documentation-only and does not affect performance.
• Concrete fix: None required.

Maintainability

Finding: TWFE dedup item is now recorded as waived/deferred.

• Severity: P3 informational
• Impact: Acceptable. The row is now under Won't-fix / waived, and the stated premise is supported by existing shared helper usage and tests.
• Concrete fix: None required.

Reference: TODO.md:L146; helper coverage in tests/test_utils.py:L2084-L2127.

Tech Debt

Finding: Deferred work is properly tracked.

• Severity: P3 informational
• Impact: This is exactly the accepted deferred-work pattern: a prior actionable follow-up is moved into documented waived debt with rationale and PR/date attribution.
• Concrete fix: None required.

Security

Finding: None.

• Severity: None
• Impact: TODO-only diff; no secrets, credentials, runtime inputs, or security-sensitive code paths changed.
• Concrete fix: None required.

Documentation/Tests

Finding: No new tests added for the TODO move.

• Severity: P3 informational
• Impact: Acceptable for docs-only bookkeeping. Existing tests already cover the shared dummy builder and TWFE/DiD full-dummy consistency, including survey variants at tests/test_estimators_vcov_type.py:L1073-L1198.
• Concrete fix: None required.