Add per-slave ESP-NOW OTA progress over UART and fix dashboard updates.

Expose OTA_SLAVE_PROGRESS on the master, track per-slave state during
distribution, run ESP-NOW OTA in a background task so the host can poll
while slaves update, and show master/slave progress in the dashboard
with table layout and faster WebSocket refresh during uploads.

Co-authored-by: Cursor <cursoragent@cursor.com>

main/proto/uart_messages.proto

@@ -18,6 +18,7 @@ enum MessageType {
   OTA_END = 18;
   OTA_STATUS = 19;
   OTA_START_ESPNOW = 20;
+  OTA_SLAVE_PROGRESS = 21;
 }
 
 message UartMessage {
@@ -36,6 +37,8 @@ message UartMessage {
     AccelDeadzoneResponse accel_deadzone_response = 12;
     EspNowUnicastTestRequest espnow_unicast_test_request = 13;
     EspNowUnicastTestResponse espnow_unicast_test_response = 14;
+    OtaSlaveProgressRequest ota_slave_progress_request = 15;
+    OtaSlaveProgressResponse ota_slave_progress_response = 16;
   }
 }
 
@@ -119,10 +122,32 @@ message OtaPayload {
 message OtaEndPayload {}
 
 // Device → host status (also used as ACK after each 4 KiB written).
-// status: 1=preparing, 2=ready, 3=block_ack, 4=success, 5=failed
+// status: 1=preparing, 2=ready, 3=block_ack, 4=success, 5=failed, 6=distributing
 message OtaStatusPayload {
   uint32 status = 1;
   uint32 bytes_written = 2;
   uint32 target_slot = 3;
   uint32 error = 4;
 }
+
+// Host → master: query ESP-NOW slave OTA progress (client_id 0 = all slaves in session).
+message OtaSlaveProgressRequest {
+  uint32 client_id = 1;
+}
+
+message OtaSlaveProgressEntry {
+  uint32 client_id = 1;
+  uint32 bytes_written = 2;
+  uint32 total_bytes = 3;
+  /** 0=idle, 1=preparing, 2=ready, 3=distributing, 4=success, 5=failed */
+  uint32 status = 4;
+  uint32 error = 5;
+}
+
+message OtaSlaveProgressResponse {
+  bool active = 1;
+  uint32 total_bytes = 2;
+  uint32 aggregate_bytes = 3;
+  uint32 slave_count = 4;
+  repeated OtaSlaveProgressEntry slaves = 5 [(nanopb).max_count = 16];
+}