Reduce idle work and disk churn with native resource diagnostics#2679

fix(observability): close lifecycle resource gaps

fix(settings): preserve advanced activity overrides

fix(server): bound telemetry supervisor failures

fix(server): preserve telemetry error structure

fix(desktop): decode packaged host telemetry

fix(desktop): honor backend stop deadlines

Address remaining observability review findings

Address observability review findings

Polish resource diagnostics and harden RPC scopes

Reduce provider observability disk churn

Remove legacy local backend and telemetry RPC

Buffer backend output before persisting failure logs

Route desktop telemetry control and stale power handling

feat: add native resource telemetry

Add background activity policy and host power monitoring