Description

In #4695 / #4637 we introduced some waiting logic for regular operations. One of these waits was related to the log retrieval.
Although this seems like a good idea, it doesn't fit all purposes. Some Pods might have started but aren't ready, or might have failed which makes them unready. However, they might have logged something which can be crucial to detect bugs and so on.

For all of these use cases, we should remove the Readiness check and replace by something else to simply detect if the Pod is live and can be queried for logs.

/cc @shawkins