Share one helper for the HHS state median income LIHEAP limit #10005

Problem

Four state LIHEAP income-limit formulas compute the same HHS state median income (SMI) limit, each with its own copy of the household-size adjustment and rounding:

Each formula floors the state's rate (60%) times the four-person SMI, multiplies by the size factor, and floors again. That order reproduces the published limits: all 20 Mississippi rows, Georgia's 16 FY2026 limits, Florida's FY2026 sizes 1-12, and every HHS FFY2025 and FFY2026 Ohio amount. The size factor also repeats the calculation inside smi() in policyengine_us/variables/gov/hhs/hhs_smi.py.

HHS describes the calculation in the footnote to its SMI table: "60 percent of each state's estimated median income for a four-person family is multiplied by the following percentages to adjust for household size for LIHEAP" (LIHEAP IM 2025-02, Attachment 4). The footnote does not state a rounding order; the floor-then-floor order is a reconciliation against the printed tables.

Proposed change

• Add a size-factor helper to hhs_smi.py, for example smi_size_factor(unit_size, period, parameters), and have smi() call it.
• Add a helper that returns the LIHEAP SMI limit: floor the rate times the four-person SMI, multiply by the size factor, floor again. Let the caller pass the date for the SMI lookup, since Ohio reads the prior year's SMI from the 2026-27 season (smi_year_lag).
• Move Mississippi, Ohio, Georgia and Florida onto the helper. Florida's FY2025 branch, which rounds the unfloored amount, keeps its own path.
• co_ccap_smi.py also reads household_size_adjustment directly and can use the size-factor helper.

Acceptance criteria

• Every existing test expectation in the four programs stays the same.
• No new parameters.
• Ohio, Georgia and Florida move onto the helper after their PRs merge.