fix(admin): explain Trino recovery blockers

[benben]

Summary

Keep the Trino recovery action visible when an instance is not eligible, and explain the exact blockers instead of hiding the confirmation form.

• Show actual stored serving and minimum counts, non-draining phase, frozen pool, and missing or invalid admitted identity fields.
• Keep the new-request action disabled for pending, recorded, retired, or unavailable-preview states, alongside the relevant status.
• Preserve existing authorization, exact-identity checks, capacity floor, immutable requests, and explicit confirmation requirements. No API, operator, database, or deployment changes.
• Document that draining alone does not imply recovery eligibility; insufficient serving capacity must be restored first.

Validation

• Three regression tests reproduced the missing eligibility feedback before implementation.
• just ui-test: 258 tests pass, including confirmation reset after capacity recovers; typecheck and production build pass.
• git diff --check: clean.
• just lint was run and still reports three unrelated baseline issues: cached rawConn.Close errcheck and two reflect.Ptr warnings in installed node_modules/flatted Go source. This PR changes no Go files.

This UI change does not bypass lifecycle checks or resolve a pool capacity deficit. No live recovery is part of this PR.

[github-actions]

Test Impact Plan

Deterministic summary of how this PR changes tests, CI runners, and coverage-risk signals.

Summary

Area	Added	Changed	Deleted
Test files	0	0	0
E2E/journey files	0	0	0
Workflow files	0	0	0

Signals

• Test cases: +0 / -0
• Assertions: +0 / -0
• Skips or known failures added: 0
• Workflow continue-on-error added: 0
• Workflow path filters added: 0
• Test commands removed from justfile: 0
• E2E/journey retry lines added: 0

Coverage risk: neutral or increased

No coverage-reduction warnings detected.