If you register it by the virtual hostname, that's actually tied to a VIP.
The virtual hostname and VIP are both included as cluster resources, so when a fail over occurs, monitoring will follow the virtual hostname which will still resolve to the VIP which will now be bound to the new node.