snake_case sweep: rename camelCase/Hungarian leftovers in test toolchain

[sbryngelson]

The Python toolchain uses snake_case (repo convention), but the test runner still has camelCase and Hungarian-notation leftovers. All are local variables / function parameters → low-risk mechanical rename, no public-API or golden-file impact.

toolchain/mfc/test/case.py (define_case_d, L442-454):

• newTrace → new_trace (L442, 450, 451, 454)
• newMods → new_mods (L442, 448)

toolchain/mfc/test/test.py:

• bFoundFrom, bFoundTo → found_from, found_to (L92, 97, 99, 102, 105) — Hungarian b prefix
• nAttempts → n_attempts (L606, 613, 622, 631, 659, 662) — Hungarian n prefix
• (from_i at L93/96/100 is already snake_case — leave it.)

toolchain/mfc/test/cases.py (pervasive, ~80 hits):

• dimInfo → dim_info, dimCmp → dim_cmp, dimParams → dim_params (throughout the alter_* functions / get_dimensions / foreach_dimension)
• patchID → patch_id
• casesToSkip → cases_to_skip (L1730, 1775)
• caseSize → case_size (L1793, 1794, 1800)

dimInfo/dimCmp/dimParams/patchID recur dozens of times across the alter_* helpers (several are nested closures over dimInfo), so a careful scoped rename per-function is needed, but it remains purely mechanical. Run ./mfc.sh lint (ruff) after to confirm the suite still imports cleanly. File as one combined issue.

---

_{Filed from a repo-wide code-cleanliness review; verified against master @ 40dde5e.}

Code references

• toolchain/mfc/test/case.py:442-454 — newTrace / newMods
• toolchain/mfc/test/test.py:92-105 — bFoundFrom / bFoundTo
• toolchain/mfc/test/test.py:606-631 — nAttempts
• toolchain/mfc/test/cases.py:185-258 — dimInfo / patchID