Out of the Kubernetes docs a kubectl tool has "three kinds of object management":
While the first and the last options' use cases are more or less clear, the second one really makes me confusing.
Moreover in the concepts section there is a clear distinction of use cases:
Also imperative style is recommended for CKA certification so it seems to be preferred for day-to-day cluster management activities.
But once again what is a best use case / practice for "Imperative object configuration" option and what is the root idea behind it?
There are two basic ways to deploy to Kubernetes: imperatively, with kubectl commands, or declaratively, by writing manifests and using kubectl apply. A Kubernetes object should be managed using only one technique. It is better to use only one way for the same object, mixing techniques for the same object results in undefined behavior.
Imperative object configuration creates, updates and deletes objects using configuration files, which contains fully-defined object definitions. You can store object configuration files in source control systems and audit changes more easily than with imperative commands.
You can run kubectl apply, delete, and replace operations with configuration files or directories containing configuration files.
Please refer to official documentation, where everything is fully described with examples. I hope it is helpful.