Windows Failover Clustering is one of many ways on Windows to provide high-availability. In general, you would only use it for applications that don’t provide their own high-availability solution. I only recommend Failover Clustering to be used in cases where you can only run one instance of your app (typically a legacy app), and you run it inside a Hyper-V VM and then Hyper-V Clustering can failover that VM to a new physical host. It’s still running the same VM and app instance, by using shared iSCSI/SAN/SMB storage to hold the VM image.
I’ve never seen that setup in a Docker setup because in a container world, you’re usually moving toward a model where you obtain high-availability by running many containers from the same app image, and your data stores are replicated/mirrored across containers as well. “container failover” for Swarm really means “start a new container from the same image as the container that failed”, which is nothing like what happens in Windows Failover Clustering.
My advice, avoid Windows Failover Clustering in your Docker/Swarm rollout. Combining it with Docker is an anti-patern.