[Failing Test] cmd/plugin `TestPreloadCommand` times out after 10m0s on OCI-override subtests

### Which jobs are failing?

periodic-cluster-api-operator-test-main
https://prow.k8s.io/view/gs/kubernetes-ci-logs/logs/periodic-cluster-api-operator-test-main/2083657800268910592
https://prow.k8s.io/view/gs/kubernetes-ci-logs/logs/periodic-cluster-api-operator-test-main/2084020193507414016
https://prow.k8s.io/view/gs/kubernetes-ci-logs/logs/periodic-cluster-api-operator-test-main/2084382586594922496

### Which tests are failing?

sigs.k8s.io/cluster-api-operator/cmd/plugin/cmd TestPreloadCommand
sigs.k8s.io/cluster-api-operator/cmd/plugin/cmd TestPreloadCommand/multiple_providers_with_OCI_override

### Since when has it been failing?

Aug 1 2026, and on every daily run since (Aug 1, Aug 2, Aug 3 all red as of filing).
Last known-green run was Jul 31 2026: on the exact same commit `46918b6`
(https://github.com/kubernetes-sigs/cluster-api-operator/tree/46918b657109212410b71c7e0401f63b0198406d, PR #1093) as the first failing run on Aug 1.

### Testgrid link

https://storage.googleapis.com/k8s-triage/index.html?job=periodic-cluster-api-operator-test-main&text=TestPreloadCommand

### Reason for failure (if possible)

```
=== Failed
=== FAIL: cmd/plugin/cmd (0.00s)
Creating new test environment
panic: test timed out after 10m0s
running tests:
TestPreloadCommand (9m53s)
TestPreloadCommand/multiple_providers_with_OCI_override (4m27s)
...

=== FAIL: cmd/plugin/cmd TestPreloadCommand/multiple_providers_with_OCI_override (unknown)
Added file: core-cluster-api-v1.10.0-beta.0-components.yaml
Added file: core-cluster-api-v1.10.0-beta.0-metadata.yaml
Added file: infrastructure-docker-v1.10.0-beta.0-components.yaml
Added file: infrastructure-docker-v1.10.0-beta.0-metadata.yaml
Packaged manifests
```

Test hangs right after logging "Packaged manifests" (the OCI-override preload/package step), then Go's test binary is killed by the default 10-minute test timeout. The goroutine dump shows the process parked waiting on a client-go informer to sync, this looks like a hang/deadlock somewhere in the OCI-override packaging path, not an application panic.

### Anything else we need to know?

- Identical failure signature and identical two subtests fail on all 3 observed runs.
- Each occurrence burns a full 10-minute CI timeout slot.

### Label(s) to be applied

/kind failing-test
/area testing

Start with cmd/plugin/cmd TestPreloadCommand, especially the multiple_providers_with_OCI_override subtest, and reproduce the timeout using the linked periodic job logs. Inspect the goroutine dump and the OCI-override preload/package step after "Packaged manifests". Done means the subtest completes without waiting for informer sync and the full TestPreloadCommand passes.

Written by the indexing model from the issue text.