Please use your available web-tests-analyzer MCP tools to analyze the performance differences between these dataset groups:

Group 1 (Baseline):
- <dataset path or gs:// URI>
- <optional additional metrics files>

Group 2 (Experiment):
- <dataset path or gs:// URI>
- <optional additional metrics files>

<Additional groups>

I would like you to write a comprehensive, beautifully formatted Markdown report comparing their performance. Please follow this workflow:

CRITICAL RULE: DO NOT write any custom Python or Node scripts (like `summarize.py`) to parse the JSON files manually. ALWAYS use the provided MCP tools, which inherently support statistical permutation tests, Holm-Bonferroni correction, and polarity logic.

Environment Setup: Use get_dataset_metadata to extract the exact OS, browser, and hardware metadata for all datasets.

Coverage Check: Use compare_datasets to confirm which test suites they share in common.

The Highlight Reel: Iterate through the common tests from step 2 and use get_significant_changes for each specific test to pull a summary of the most critical, statistically significant regressions and improvements. (Rely on the default importantMetrics filter to keep the signal-to-noise ratio high).

Deep Dives: If you notice any particularly severe regressions or interesting anomalies in a specific test suite from step 3, use analyze_test to do a deeper dive into that specific suite to pull more context.

Report Formatting: Please structure the final report with:

An Executive Summary highlighting the overarching narrative.
A Test Environment section detailing the metadata.
A Detailed Breakdown of the significant metric changes, utilizing Markdown tables to cleanly compare the group means along with their percentage change.