fix(health-monitoring): resolve HealthMonitoring-004,006,010,011,012 — heartbeat-doc accuracy, testable sequence seeding, logged failures, dead-code removal

src/ScadaLink.HealthMonitoring/CentralHealthAggregator.cs

@@ -210,10 +210,12 @@ public class CentralHealthAggregator : BackgroundService, ICentralHealthAggregat
             var state = kvp.Value;
             if (!state.IsOnline) continue;
 
-            // Use LastHeartbeatAt — heartbeats arrive frequently from any
+            // Use LastHeartbeatAt — heartbeats arrive every ~5s from any
             // healthy site node (cadence owned by Cluster Infrastructure /
-            // SiteCommunicationActor), so OfflineTimeout only fires when no
-            // node can reach central, not during single-node failovers.
+            // SiteCommunicationActor — CommunicationOptions.TransportHeartbeatInterval),
+            // so the 60s OfflineTimeout tolerates several missed heartbeats and
+            // only fires when no node can reach central, not during single-node
+            // failovers.
             //
             // The synthetic "central" site has no heartbeat source — its only
             // signal is the 30s CentralHealthReportLoop self-report — so it gets