update gcr.io/kubernetes-ci-images references to gcr.io/k8s-staging-ci-images

[CecileRobertMichon]

The gcr.io/kubernetes-ci-images was removed, we should replace all references in CAPZ scripts for CI to gcr.io/k8s-staging-ci-images.

There might also be references in the CAPI test framework, we should fix those too.

The error we're observing in CI:

[   79.681413] cloud-init[1825]: [2021-07-12 14:43:01] 	[ERROR ImagePull]: failed to pull image gcr.io/kubernetes-ci-images/kube-controller-manager:v1.22.0-beta.1.136_8ab66dcc4e481a: output: time="2021-07-12T14:42:57Z" level=fatal msg="pulling image failed: rpc error: code = NotFound desc = failed to pull and unpack image \"gcr.io/kubernetes-ci-images/kube-controller-manager:v1.22.0-beta.1.136_8ab66dcc4e481a\": failed to resolve reference \"gcr.io/kubernetes-ci-images/kube-controller-manager:v1.22.0-beta.1.136_8ab66dcc4e481a\": gcr.io/kubernetes-ci-images/kube-controller-manager:v1.22.0-beta.1.136_8ab66dcc4e481a: not found"

/assign @jackfrancis

[CecileRobertMichon]

related to kubernetes/k8s.io#2318