Ensure /join response for Pending nodes is consistent

[eddyashton]

Noticed while debugging something else:

2024-11-19T11:32:39.363335Z -0.001 0   [info ] ../src/node/node_state.h:623         | !!! Received join response:
{
  "node_id": "80a95a1c87127f0c9351df36cba2d70bca6f85f05bd15e09493c6c17ea27c908",
  "node_status": "Pending"
}

2024-11-19T11:32:39.365522Z -0.002 0   [info ] ../src/node/node_state.h:623         | !!! Received join response:
{
  "node_status": "Pending"
}

I think this is harmless, but may bite us in future. Should refactor the code to ensure the response is the same, whether the node is newly added or already known to be pending.