VM Monitoring uses a heartbeat between the host and the VM through VMware Tools that are installed on the VM.
It's mainly there to detect a crash in the VM, like a Windows BSoD, or something else that causes the host to no longer receive heartbeat replies from the VMware Tools driver in the VM.
I'm afraid that your test is not of the proper type to check if VMware HA events or VM Monitoring are working as designed.
To test that VM Monitoring is working, you will have to figure out some way to cause a BSoD crash inside your VM, and then observe the behavior.
To test if VMware HA is working, you would have to sever all network connections from that host to other hosts in the cluster, and (if you are using datastore heartbeats as well) sever all storage paths to that host too.
More info on VM Monitoring (still works this way on current versions):
Also, you should probably change the Host Isolation Response to either "Power Off" or "Shut Down" from the default choice (Leave Powered On).