Provide failure, flaky test, and regression-focused summaries

Context

Maintainers running large test suites generally do not need to inspect successful results individually. User feedback highlighted workflows where thousands of tests run regularly, while the information requiring attention is primarily failures, flaky tests, errors, and newly introduced regressions.

Displaying large amounts of successful results makes it harder to identify actionable information.

Goal

Provide concise summaries focused on results that require developer attention. A maintainer should be able to quickly understand what failed, what appears flaky, and what changed compared with previous runs without navigating through large sets of successful results.

Points to discuss
• What should appear in the default summary.
• Distinction between failure, error, and flaky results.
• How flaky behavior should be detected and represented.
• Grouping repeated failures across environments.
• Integration with A/B Comparison.
• Whether summaries should be reusable by both the dashboard and reporting/notification mechanisms.

Start with the issue's open questions about default summaries, failure/error/flaky distinctions, repeated failures across environments, and A/B Comparison. Review how the dashboard and reporting/notification mechanisms currently present test results; done means an agreed summary design that identifies failures, flaky tests, and regressions without requiring successful results to be inspected individually.