I guess if there is a rule as you say, it must be for certain specific case. In general there is no such rule and culturing and subculturing would purely be based on you purpose and aim.

As long as you have a backup culture there should be no problem. A potential reason for that rule is that the lingering antibiotic might reduce viability when exposed to the different media. If you grow your bacteria in an overnight culture without selection it is no issue.

Of course they "can" - assuming the selective agent doesn't inhibit. What is the purpose of your subculture ? You should work with from stock culture and that should not be on selective medium. If isolated from a potentially complex population, pure culture should be finally established on nonselective media to ensure no nontarget microbe managed to get through the selective media .

Thanks for your thoughts and hints. I think one of the backgrounds behind it might be that often our pure cultures and strains obtained from plants are meant to go into a pathogenicity assay, and a passage over several (different) selective media containing antibiotics or other chemicals might change its virulence, or its phenotype in general. Selective pressure on the colonies obtained after spreading it on and subculturing it from a selective medium might select for colonies that differ in certain traits. Let me know if you consider this erroneous. Thanks!

In the process of isolation (obtention of pure cultures), one usually starts by plating out the original sample onto an adequate selective agar medium and then selecting one or several well isolated colonies with the desired characteristics. When selecting an isolated colony, we presume that we have a pure culture (i.e., a population that descends from a single bacterial cell), but that is not always the case, due to several reasons. If we further streak this colony onto a selective agar again, eventual contaminant bacteria present in the original colony will be inhibited, but may persist. On the other hand, if you use a non-selective medium, contaminants will show up as different colonies, and you may select only the colonies that have the important characteristics you're looking for, leaving behind contaminants.

Furthermore, passing bacteria through a selective medium will stress them and they will not be in their best physiological shape for further testing.

A note of caution: you should not subculture bacterial isolates multiple times, once you estabished your pure culture, and that applies to either a selective or a non-selective culture medium. You'll cause mutations and alter their properties. I would advise you to constitute a stock culture of each isolate (in slants, for instance) and always go back to that same stock to perform each test on selective agars.