Address P2/P3: REGISTRY replicate matrix, ROADMAP qualifier, soften wording

- Update REGISTRY.md replicate-weight support matrix: CS now supports
  covariates with replicate weights (IF-based path is covariate-agnostic,
  shipped in Phase 7a)
- Qualify ROADMAP.md: "replicate weights supported for 12 of 15" instead
  of "across all estimators"
- Soften consolidated deferred section from "single source of truth" to
  "summary of major remaining limitations" with TODO.md cross-reference

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: igerber

ROADMAP.md

@@ -16,7 +16,7 @@ diff-diff v2.8.4 is a **production-ready** DiD library with feature parity with
 - **Sensitivity analysis**: Honest DiD (Rambachan-Roth), Pre-trends power analysis (Roth 2022)
 - **Study design**: Power analysis tools
 - **Data utilities**: Real-world datasets (Card-Krueger, Castle Doctrine, Divorce Laws, MPDTA), DGP functions for all supported designs
-- **Survey support**: Full `SurveyDesign` with strata, PSU, FPC, weight types, replicate weights (BRR/Fay/JK1/JKn/SDR), Taylor linearization, DEFF diagnostics, subpopulation analysis — integrated across all estimators (see [survey-roadmap.md](docs/survey-roadmap.md))
+- **Survey support**: Full `SurveyDesign` with strata, PSU, FPC, weight types, replicate weights (BRR/Fay/JK1/JKn/SDR), Taylor linearization, DEFF diagnostics, subpopulation analysis — integrated across all estimators (replicate weights supported for 12 of 15; see [choosing_estimator.rst](docs/choosing_estimator.rst#survey-design-support) for the compatibility matrix)
 - **Performance**: Optional Rust backend for accelerated computation; faster than R at scale (see [CHANGELOG.md](CHANGELOG.md) for benchmarks)
 
 ---

docs/methodology/REGISTRY.md

@@ -2307,9 +2307,10 @@ variance from the distribution of replicate estimates.
   design structure is fixed and dropped replicates contribute zero to the
   sum without changing the scale. Survey df uses `n_valid - 1` for
   t-based inference.
-- **Note:** Replicate-weight support matrix:
-  - **Supported**: CallawaySantAnna (reg/ipw/dr without covariates, no
-    bootstrap), ContinuousDiD (no bootstrap), EfficientDiD (no bootstrap),
+- **Note:** Replicate-weight support matrix (12 of 15 estimators):
+  - **Supported**: CallawaySantAnna (reg/ipw/dr with or without covariates,
+    no bootstrap; IF-based replicate variance is covariate-agnostic),
+    ContinuousDiD (no bootstrap), EfficientDiD (no bootstrap),
     TripleDifference (all methods), LinearRegression (OLS path),
     DifferenceInDifferences (no-absorb via LinearRegression dispatch,
     absorb via estimator-level refit), MultiPeriodDiD (no-absorb via

docs/survey-roadmap.md

@@ -275,8 +275,9 @@ deferred to the consolidated list below:
 
 All items below raise an error when attempted (`NotImplementedError` or
 `ValueError` depending on the estimator), with a message describing the
-limitation. This is the single source of truth for remaining survey
-limitations.
+limitation. This is a summary of the major remaining survey limitations.
+See also `TODO.md` for general tech debt items (e.g., multi-absorb +
+survey weights).
 
 ### Replicate Weights Not Supported