To try keep this short:
Is this a widely known issue? I have hit it a few times when attempting to test the nsfarmserverurl failover, but I'm not sure whether it's a misconfiguration or a genuine bug.
Is there any way of debugging the "failover" when one of the farm servers is down? I have tried every different debugging level, but still see nothing helpful in the error log on the server with the link configured. For example I would like to see which backend servers it is attempting to reach at what time. (To trigger the chain I'm attempting a password change on a client, which simply hangs when the first farm server is down)
I can provide information to reproduce if necessary.