Alright - went to the 2nd control node in /etc/systemd/system/k3s.service and updated the command line args to specify the 3rd control node with --server and that node is now healthy

aaaaand I went to the 1st control node, added the --token and --server lines and now it's healthy

Props to the kubernetes/etcd/k3s devs. I thought that cluster was smoking charred remains

Anyway -- LMK if there are other resources I could have / should have looked at first. I'm curious what others would have done to recover the cluster