feat(telemetry): add TX Processing Latency by Type panel to dashboard

Shows p95 latency of tx.process span broken down by tx_type. Works for
both received and locally-processed transactions, unlike the tx.transactor
panel which requires the node to be synced and applying.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

docker/telemetry/grafana/dashboards/transaction-overview.json

@@ -495,6 +495,50 @@
         },
         "overrides": []
       }
+    },
+    {
+      "title": "TX Processing Latency by Type (p95)",
+      "description": "Per-transaction-type processing latency (tx.process span duration). Shows which transaction types take longest to process end-to-end.",
+      "type": "timeseries",
+      "gridPos": {
+        "h": 8,
+        "w": 24,
+        "x": 0,
+        "y": 48
+      },
+      "options": {
+        "tooltip": {
+          "mode": "multi",
+          "sort": "desc"
+        },
+        "legend": {
+          "displayMode": "table",
+          "placement": "right",
+          "calcs": ["mean", "max", "lastNotNull"]
+        }
+      },
+      "targets": [
+        {
+          "datasource": {
+            "type": "prometheus"
+          },
+          "expr": "histogram_quantile(0.95, sum by (le, tx_type) (rate(traces_span_metrics_duration_milliseconds_bucket{exported_instance=~\"$node\", span_name=\"tx.process\", tx_type!=\"\"}[5m])))",
+          "legendFormat": "p95 {{tx_type}}"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": {
+            "axisLabel": "Latency (ms)",
+            "spanNulls": true,
+            "insertNulls": false,
+            "showPoints": "auto",
+            "pointSize": 3
+          }
+        },
+        "overrides": []
+      }
     }
   ],
   "schemaVersion": 39,