list: test the Bash fast path against the Ruby command

[aholland]

The follow-up promised on #23235: bare brew list invocations are served by the Bash fast path in production, while the test suite's examples enter at the Ruby command — so a feature added to one implementation can pass its own tests while production serves the other. That is exactly how #23235's warning shipped inert on its main path.

This adds a parity test: both implementations run against the same installed state (a formula, a cask, and a broken Caskroom symlink) across the flag variants both serve — bare, -1, --formula, --cask — and their stdout, stderr and exit status are diffed against each other, not against hand-written expectations. Drift now fails in CI whichever side changed. (--versions --json is excluded: the Ruby command intentionally defers that to the Bash path.)

Red/green: disabling the Bash-side warning from #23245 — recreating the drift that actually shipped — fails the test.

---

• Have you followed our Contributing guidelines?
• Have you checked for other open Pull Requests for the same change?
• Have you explained what your changes do? Performance claims (e.g. "this is faster") must include Hyperfine benchmarks.
• Have you explained why you'd like these changes included, not just what they do?
• For bug fixes, have you given step-by-step brew commands to reproduce the bug?
• Have you written new tests (excluding integration tests)? Here's an example.
• Have you successfully run brew lgtm (style, typechecking and tests) locally?

---

• AI was used to generate or assist with generating this PR.

Written with Claude in Claude Code under my direction and review. Red/green verified as described above. brew typecheck, brew style and changed tests clean locally.

[MikeMcQuaid]

Thanks!