Move taxable-income ceiling from SALT to aggregate itemized deductions

[saisharan0103]

Issue #9167 identifies that salt_deduction diverges from the §164(b)(7) SALT amount when AGI minus exemptions is below otherwise allowable SALT.

Root cause: the simulation-only taxable-income ceiling was applied inside salt_deduction.formula, so the component variable reflected a whole-return itemized deduction constraint instead of the statutory SALT deduction. That also made gov.simulation.limit_itemized_deductions_to_taxable_income affect only SALT rather than aggregate itemized deductions.

Changes:

• Remove the simulation ceiling branch from salt_deduction.formula, leaving the variable to return min_(salt_cap, salt).
• Apply the simulation ceiling in itemized_taxable_income_deductions.formula after itemized_taxable_income_deductions_reduction, limiting aggregate itemized deductions to max_(0, adjusted_gross_income - exemptions) when the flag is true.
• Add a focused 2026 low-AGI SALT regression case.
• Include the required changelog fragment.

[codecov]

Codecov Report

❌ Patch coverage is 85.71429% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 96.66%. Comparing base (838be3f) to head (18cacca).
⚠️ Report is 14 commits behind head on main.

Files with missing lines	Patch %	Lines
...ns/itemizing/itemized_taxable_income_deductions.py	85.71%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##              main    #9191      +/-   ##
===========================================
- Coverage   100.00%   96.66%   -3.34%     
===========================================
  Files            2        2              
  Lines           31       30       -1     
  Branches         0        1       +1     
===========================================
- Hits            31       29       -2     
- Partials         0        1       +1     

Flag	Coverage Δ
unittests	96.66% <85.71%> (-3.34%)	⬇️

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

☔ View full report in Codecov by Harness.
📢 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]

Cross-family dual review (Claude Fable 5 + GPT-5.6-sol, independent). The federal intent matches #9167 — salt_deduction now reports the §164(b)(7) amount and the simulation-only ceiling moves to the aggregate — and the new fixture's arithmetic is correct. Blocked on blast radius:

1. Direct state consumers change materially. AZ (az_itemized_deductions), CO (co_state_addback), and UT (ut_federal_deductions_for_taxpayer_credit) read salt_deduction and previously observed the AGI-capped component. Differential sim against this PR's parent: 2026 CO single, AGI $20,000, SALT $30,000, state withholding $25,000 → co_state_addback $3,900 → $13,900, co_income_tax $152.60 → $592.60, with federal taxable income unchanged. Each of those states needs a capped-vs-uncapped ruling from its own worksheet (CO DR 0104 line 5e treatment, AZ Form 140 Schedule A adjustments, Utah §59-10-1018) plus tests — or the consumers need insulating.
2. The relocated ceiling now caps all itemized deductions, not just SALT. A household with zero SALT and $30,000 of other itemized deductions against AGI − exemptions = $10,000 drops from $30,000 to $10,000 of reported deductions, which can flip tax_unit_itemizes and feeds QBID/AMT and many state starting points. That's a much wider behavior change than the issue's low-AGI SALT reporting defect, with no neutrality evidence or coverage (the single new fixture has SALT only and no state).