Add diagnostics for defaulted health-status inputs used in Medicaid eligibility

## Type

Mixed rules and data.

## Context

Several Medicaid and health-coverage rules depend on input leaves that may be unavailable or default to false in broad survey/population datasets. Examples include medically frail or special medical needs status, recent incarceration status for Medicaid continuous eligibility, emergency medical condition, long-term care service receipt, and related categorical-exemption inputs.

These inputs are legally meaningful, but in a Populace-style national dataset the difference between “observed false” and “not available, defaulted false” matters.

## Why this matters

Defaulting eligibility-sensitive statuses to false can understate eligibility or fail to model protected/exception populations. The model should make those assumptions visible so household-level results and aggregate budget estimates can be interpreted correctly.

## Suggested direction

- Identify Medicaid/health variables where missing data are currently represented as false but the status materially affects eligibility or cost.
- Add diagnostics or companion variables that quantify records relying on defaulted assumptions.
- Document default assumptions in variable docs or tests.
- Where useful, add explicit `*_unknown` or imputation-ready inputs rather than overloading false.

## Acceptance criteria

- Eligibility-sensitive health-status inputs have documented default semantics.
- Tests or diagnostics quantify the population affected by default-false assumptions.
- Populace can distinguish observed false values from unavailable values for high-impact statuses when source data support that distinction.
- Model users can see which Medicaid/health pathways are limited by missing status data.