During precaching, cluster status entry gets missing if the cluster is unavailable temporarily while TALM reconciles

4.12.0

1. Create a CGU with precaching enabled
2. While waiting for precaching to complete, make one of the clusters no longer match the cluster selector
3. After at least one reconcile, make the cluster match the selector again.

The precaching status for that cluster gets lost and the CGU is stuck with failing to process that cluster, as empty state is not handled

The cluster status should not disappear and precaching processing for that cluster should resume and complete successfully once it's selectable again.