I have configured the windows task scheduler with any node type using powershell commands in the cluster as Active/Passive.
If one system goes down, the task scheduler is switching to another node and runs from that node when schedule time triggers.
But if the task is running on one node, and before task completion, if the system goes down, it is not re-starting in a new node, though I have a configured the system to re-execute the task if it fails.