Many of the internal aspects of the host are monitored.  However most failures impacting the compute aspects of the host occur rapidly and cannot be predicted in a time frame where that knowledge is actionable.

As for migrating guests from degraded hosts, many of the reasons for a degraded state, like a UPS, switch, or storage failure will also impact a constellation of other hosts or may be transient failure like a switch rebooting which will resolve itself.   So do you migrate immediately?   Do you wait for a few minutes before you declare the host nonviable?  depends on the failure and the infrastructure.

interested !