feat: add include_failed_runs benchmark override

[edmundmiller]

Summary

• add an include_failed_runs / include-failed-runs override for benchmark report generation
• keep the current default behavior unchanged: failed workflows remain excluded unless the flag is set
• continue excluding cancelled workflows
• plumb the flag through the Python CLIs, Nextflow params, schema, and workflow wiring
• add regression coverage for the new override behavior

Why

When comparing benchmark runs, it is useful to inspect workflow-level failed runs in the downstream HTML sections and charts instead of only seeing them dimmed/excluded in the run summary.

Test plan

• UV_NO_CONFIG=1 uv run --with typer --with pyyaml --with jinja2 --with pyarrow --with pytest --with httpx pytest -q modules/local/aggregate_benchmark_report_data/tests/test_aggregate.py::test_failed_and_cancelled_runs_only_appear_in_overview modules/local/aggregate_benchmark_report_data/tests/test_aggregate.py::test_include_failed_runs_override_includes_failed_in_downstream_sections
• UV_NO_CONFIG=1 uv run --with typer --with pyyaml --with jinja2 --with pyarrow --with pytest --with httpx pytest -q modules/local/aggregate_benchmark_report_data/tests/test_aggregate.py modules/local/render_benchmark_report/tests/test_render.py
• regenerated a local HTML report with failed runs included and verified the failed run IDs appear in downstream sections

Notes

• this change does not address the separate latest-main regressions around raw tw runs dump tarballs or the aggregate-stage symlink/import issue
• this PR focuses only on the include-failed-runs override

[FloWuenne]

@copilot resolve the merge conflicts in this pull request

[Copilot]

@copilot resolve the merge conflicts in this pull request

Resolved and pushed. I merged origin/main into this branch, resolved the conflicts in README.md, nextflow.config, and nextflow_schema.json, and preserved both feature sets. Commit: 565e9e3.