Refactor health status to separate status from reason

[yrobla]

Previously, BackendHealthStatus conflated WHAT the status is with WHY the backend is in that state. For example, BackendUnauthenticated was both a status value and a reason, making it unclear whether a backend was unavailable due to authentication issues or other causes.

This improves observability by providing clear separation between the backend's availability state and the underlying cause of that state.

Related-to: #3147

This change introduces BackendHealthReason to explicitly capture WHY a backend is in its current state, while BackendHealthStatus describes WHAT the state is (healthy, degraded, unhealthy, unknown).

[codecov]

Codecov Report

❌ Patch coverage is 94.04762% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 64.87%. Comparing base (e22900f) to head (b83da2a).
⚠️ Report is 73 commits behind head on main.

Files with missing lines	Patch %	Lines
pkg/vmcp/health/checker.go	86.95%	3 Missing ⚠️
pkg/vmcp/health/status.go	95.45%	1 Missing ⚠️
pkg/workloads/manager.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3421      +/-   ##
==========================================
+ Coverage   64.75%   64.87%   +0.11%     
==========================================
  Files         382      383       +1     
  Lines       37187    37224      +37     
==========================================
+ Hits        24082    24148      +66     
+ Misses      11221    11189      -32     
- Partials     1884     1887       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[yrobla]

i will implement later when we add more integration with k8s as well