I guess that's true if you're treating containers and the environment they run in as "another layer" of stuff that you use configuration management to deploy on top of your existing automation you maybe already had prior to adopting them. I do suppose that is probably true of many users still today. There is also the use case of updating your base image or AMI and then replacing slaves/followers/your servers that run the containers, more or less eliminating the need for a job that is constantly running, or running on a schedule to keep the host system up to date. It's true that in the case of container hosting systems, that is mostly being abstracted away from you, and this post was written with that mindset as we do just that. Good comment and point of view though, I think it's still something a lot of folks struggle to wrap their heads around!