fix: false positives when checking for node missing images

[emosbaugh]

What this PR does / why we need it:

The current functionality to list missing images uses the node object which only lists the first 50 images. This creates a false positives if an image does is not in this first 50. The new functionality added leverages the existing job code that mounts the cri socket to get a comprehensive list of images.

See an example false positive below:

Have images been loaded on node ethanm-24? (y/N) y

Node ethanm-24 missing image(s) registry.k8s.io/pause:3.6 registry.k8s.io/coredns/coredns:v1.8.6 registry.k8s.io/pause:3.5 registry.k8s.io/pause:3.6 ghcr.io/projectcontour/contour:v1.23.2 haproxy:2.7.6-alpine3.17 docker.io/flannel/flannel-cni-plugin:v1.1.2 rqlite/rqlite:7.10.0 registry:2.8.1 velero/velero-restic-restore-helper:v1.9.5 gcr.io/velero-gcp/velero-restic-restore-helper:v1.9.5 velero/velero-plugin-for-aws:v1.6.0 velero/velero-plugin-for-gcp:v1.6.0 velero/velero-plugin-for-microsoft-azure:v1.6.0

Which issue(s) this PR fixes:

Fixes NONE

Special notes for your reviewer:

Steps to reproduce

Does this PR introduce a user-facing change?

Fixes an issue that could cause kURL to erroneously prompt the user to run the load-images task when upgrading clusters that have many container images.

Does this PR require documentation?

NONE

[emosbaugh]

There is a bootstrapping problem here. The command runs before the k8s-util image is available on the remote node.