Keep provider-charge assumptions consistent in subsidy and out-of-pocket estimates

In the calculator, choose Colorado, one adult, children ages 3 and 7, $30,000 annual earnings, and $1,200/month per child. Leave other inputs at their defaults.

Observed:

• Subsidy: $1,893/month.
• Copay: $107/month.
• Out of pocket: $507/month.
• Explanation attributes $400 to charges above the state's maximum rate.

A direct model calculation for those exact inputs returns $2,293 subsidy and $107 out of pocket. The apparent $400 above-rate charge comes from rounding each child's charge down to the $1,000 grid level for the subsidy, while retaining $1,200 per child in the out-of-pocket calculation.

Interpolate across the charge axis where appropriate, or consistently present the reference household's expenses and results. Do not attribute a rounding artifact to state reimbursement policy. Handle charges outside the grid explicitly. Add a regression for this example and charge-grid boundaries.

Sources: src/lib/dataLookup.ts, src/components/CalculatorResults.tsx.

Start with src/lib/dataLookup.ts around line 216 and src/components/CalculatorResults.tsx around line 52, then reproduce the Colorado household example from the issue. Trace how provider charges are selected for subsidy and out-of-pocket results. Add a regression for the example and coverage for charge-grid boundaries, including values outside the grid.