Scope orphaned-event cleanup to recent ingest window to eliminate 8-hour ES CPU spikes

[Copilot]

What changed

• Limits orphaned-event stack, project, and organization scans to events created within the last three days.
• Applies the same cutoff to cardinality queries, partitioned terms queries, and delete-by-query operations.
• Captures one cutoff per job run so all three cleanup phases use an atomic boundary.
• Uses the current Elastic.Clients.Elasticsearch and repository architecture from main.
• Adds regression coverage for stack, project, and organization orphans at and immediately before the inclusive cutoff.

Why

The cleanup job runs every eight hours. Scanning all event history caused unnecessary Elasticsearch CPU spikes even though actionable orphan cleanup is limited to the practical recent-ingest window.

Verification

• dotnet build tests/Exceptionless.Tests/Exceptionless.Tests.csproj --no-restore --nologo --verbosity:minimal
• Exceptionless.Tests --filter-class Exceptionless.Tests.Jobs.CleanupDataJobTests: 28 passed
• Exact lookback-boundary regression: passed
• Local Aspire dogfood with healthy Elasticsearch 8.19.15, Redis, API, and Jobs resources

Compatibility

No public API, WebSocket, configuration, or frontend behavior changes.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you all sign our Contributor License Agreement before we can accept your contribution.
1 out of 2 committers have signed the CLA.

✅ niemyjski
❌ Copilot
_{You have signed the CLA already but the status is still pending? Let us recheck it.}