It appears that this issue concerns the KNI static-pod CoreDNS, not the cluster CoreDNS service, so I am changing the component accordingly.
It looks like we never backported the fix to respect option 119. I've proposed it in https://github.com/openshift/machine-config-operator/pull/2904 . That should fix this.
*** This bug has been marked as a duplicate of bug 2059724 ***