Kubelet deletes pod without deleting CNI resources

i see that kubelet only checks whether containers are killed and volume mounts are cleaned before removing the pod from the apiserver but doesn’t wait for the cni to be cleaned up, is that expected and any reason for not making this as a mandatory check.

Cluster information:

Kubernetes version: master
Cloud being used: GCP
Installation method: kubeadm
Host OS: ubuntu
CNI and version:
CRI and version: