Compare paper OCP against LyoPRONTO baselines

[bernalde]

Summary

• Add Issue Compare paper-reference OCP results against LyoPRONTO baselines #30 benchmark notes comparing the paper-reference direct-transcription OCPs against current LyoPRONTO quasi-steady baselines.
• Include a clear comparison table covering drying time, active constraints, controls, and temperature behavior.
• Make the recommendation explicit: keep the paper model validation-only and route LyoPRONTO-facing adapter work through Add experimental LyoPRONTO policy OCP adapter #31.
• Add a focused regression test so the comparison table, recommendation, and Add experimental LyoPRONTO policy OCP adapter #31 follow-up remain documented.

Tests run

• python -m ruff check tests/test_benchmarks.py - passed.
• python -m ruff format --check tests/test_benchmarks.py - passed, 1 file already formatted.
• PYTHONPATH=. pytest tests/test_benchmarks.py::test_paper_ocp_notes_cover_issue30_comparison -q - passed, 1 test in 3.89s.
• PYTHONPATH=. pytest tests/test_benchmarks.py -q - passed, 18 tests in 5.48s.
• PYTHONPATH=. pytest tests/ -m "not notebook and not pyomo and not slow" -q - passed, 247 tests in 73.35s.

Notes about tests not run

• Full Pyomo, notebook, and slow tests were not run locally because this change is limited to benchmark notes plus a lightweight docs regression test. The broader fast non-Pyomo/non-notebook suite passed.
• A plain targeted pytest invocation without PYTHONPATH=. failed during collection in this shell with ModuleNotFoundError: No module named 'benchmarks'; the successful commands above set the repository root on PYTHONPATH.

Closes #30

[bernalde]

Blocking issues: none.

Nonblocking issues: none.

Questions: none.

Tests run and outcomes:

• Verified documented benchmark ranges against benchmarks/results/baseline_Tsh_3x3_summary.jsonl and benchmarks/results/baseline_Pch_3x3_summary.jsonl with jq; the documented ranges match the committed data.
• python -m ruff check tests/test_benchmarks.py passed.
• python -m ruff format --check tests/test_benchmarks.py passed.
• python -m pytest tests/test_benchmarks.py::test_paper_ocp_notes_cover_issue30_comparison -q passed, 1 test.
• python -m pytest tests/test_benchmarks.py -q passed, 18 tests.
• python -m pytest tests/ -m "not notebook and not pyomo and not slow" -q passed, 247 tests.
• python -m pytest tests/ -n auto -v -m "not notebook and not pyomo" --ignore=tests/test_pyomo_models -q passed, 262 tests.

I did not run the full Pyomo or notebook suites locally because this PR changes benchmark notes and a lightweight documentation guard test; the CI-style non-Pyomo suite passed, and the stacked PR's visible GitHub checks (Deploy docs, doctests) are green.

Merge recommendation: merge as-is.