Add at-interview health coverage rule inputs

[daphnehanse11]

Closes #7816

Summary

• add person-level has_*_health_coverage_at_interview input variables for CPS current-coverage data imported in policyengine-us-data
• add has_qualifying_non_marketplace_health_coverage_at_interview, qualifying_non_marketplace_health_coverage_type_count_at_interview, and coverage_report_model_conflict for ACA-facing reported-coverage reconciliation
• add has_marketplace_health_coverage_at_interview so Marketplace/non-Marketplace conflict checks use an explicit reported current-coverage input
• add Medicaid regression coverage for current-coverage inputs and preserve continuous-coverage behavior in the CA Medicaid immigration eligibility rule

Testing

• uv run policyengine-core test policyengine_us/tests/policy/baseline/gov/aca/eligibility/has_qualifying_non_marketplace_health_coverage_at_interview.yaml policyengine_us/tests/policy/baseline/gov/aca/eligibility/coverage_report_model_conflict.yaml policyengine_us/tests/policy/baseline/gov/hhs/medicaid/medicaid_enrolled.yaml -c policyengine_us
• make format

[baogorek]

Nice work — the mixin pattern keeps the input definitions clean, and the test coverage looks solid.

Two things worth double-checking before merge:

1. has_marketplace_health_coverage default flips from True to False. Previously every person defaulted to having marketplace coverage, which fed into is_aca_ptc_eligible. Now it defaults to False (via the reported flag). That's more correct for microsim, but in the web app where users build individual households and don't set the reported flag, this could silently zero out ACA PTC for everyone. Worth confirming that pathway still works as expected.

2. receives_medicaid was an explicit input variable (docstring says it "anchors takeup calibration"). Adding a formula that delegates to the reported flag changes it from a user-settable input to a derived value. Same question: does direct assignment still take priority over the formula in contexts that depend on it?

[MaxGhenis]

use adds

[MaxGhenis]

one variable per file

[MaxGhenis]

frame around reality not surveys. what insurance do they have at the moment?

[MaxGhenis]

this would be a formula right

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (fd8a1f8) to head (2da9303).
⚠️ Report is 46 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main     #7821   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           12        13    +1     
  Lines          174       101   -73     
  Branches         0         1    +1     
=========================================
- Hits           174       101   -73     

Flag	Coverage Δ
unittests	100.00% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[MaxGhenis]

Rechecked after the cleanup pass. The structural review comments are addressed and the targeted tests pass.