so uh it was some node was stuck waiting to be uncordoned "or something". I fixed it (maybe?) by like

kubectl delete <http://machine.cluster.x-k8s.io|machine.cluster.x-k8s.io> custom-9bbdaa956de7 -n fleet-local

and patching the finalizer out. Then all the control things were able to be uncordoned and stuff is coming back.