The pool list currently shows a "healthy" status if the pool had running tasks for 97% of the expected run time.
We should also flag any pool which has "failed" tasks, whether or not the tasks ran for their expected time. Failed tasks should always be investigated.