Guard exported PUF top-tail capital gains

[MaxGhenis]

Summary

• include long_term_capital_gains_before_response in the PUF support-clone top-tail rough-AGI guard
• treat the exported pre-response LTCG column as an alias of the integrated PUF long_term_capital_gains variable for scaling eligibility
• add a regression test matching the Guard PUF support clone top tail #166 rebuild failure: internal LTCG below the cap but exported LTCG-before-response still above the $79M open-ended AGI bin

Why

The post-#166 rebuild still failed the primary latest-local MP-vs-eCPS comparison: MP loss 1.2450774476685147 vs eCPS 0.7685371816961414. The same MD PUF support clone row remained in the top bin because the guard saw max rough AGI around $70M, while the exported H5 still had long_term_capital_gains_before_response around $95M for person 192879.

Validation

• uv run ruff check src/microplex_us/pipelines/us.py tests/pipelines/test_us.py
• uv run ruff format --check src/microplex_us/pipelines/us.py tests/pipelines/test_us.py
• uv run --extra dev python -m pytest tests/pipelines/test_us.py -k 'puf_support_clone_top_tail_guard'
• uv run --python 3.13 --extra dev --extra policyengine python -m pytest tests/pipelines/test_us.py (178 passed, 6 expected warnings)

Note: an earlier full-suite run without --extra policyengine failed only from missing optional microunit / policyengine_us dependencies; the corrected Python 3.13 + policyengine-extra run passed.