Performance: optimize /analyses.json for faster console load

Summary

The management console is bottlenecked by large /analyses.json responses. We should reduce payload size and server processing time to improve total page load.

Proposed work

• Add pagination/incremental loading support for analyses listing
• Provide a compact list response shape for UI list views
• Add conditional request support (ETag / If-None-Match, optionally Last-Modified)
• Add server-side caching for expensive analyses list generation

Acceptance criteria

• Endpoint supports paginated queries with documented defaults/limits
• Compact mode returns only list-view-required fields
• ETag works correctly (304 on unchanged content)
• p95 endpoint latency and payload transfer size are measurably reduced on realistic datasets

Context

Helm/EKS-side traffic and resilience tuning has already been added in NatLabRockies/openstudio-server-helm; this issue targets app/API-level bottlenecks that infra changes alone cannot solve.

Start by tracing the /analyses.json endpoint and measuring baseline p95 latency and response size on realistic datasets. Review how the analyses list is generated and identify the API and documentation entry points involved. Done means pagination, compact responses, conditional requests, and caching meet the stated acceptance criteria with measurable performance improvements.