Would it be useful to have more than one IP address used for monitoring connectivity on an interface? For use in load balancing or in particular fail over scenarios; just in case the machine which is at the address that is being used for monitoring goes down/has ping blocked or other; either way loses connectivity resulting in essentially a false positive regarding loss of connectivity.

So essentially the method used to check connectivity (not high latency at this stage & excuse the extreme crudity) would go from;

If no ping response from ip0, loop ping response check on this interface X times then trigger failover.

to

If no ping response on ip0 AND no ping response on ip1 on this interface, loop ping response X times then trigger failover.

Am I missing something or could this well have it's uses?