Yes, that sort of etiquette (CC'ing the PI of each involved partner in a collaboration, even for exchanges at lower level) exist in some groups. It is particularly true of recent collaborations, either because they have been established recently or because the research is still in its infancy. In both case, I suppose it is nice to help PI's stay on top of how things are progressing, both on the science and on the interpersonal relationships (like, is everyone acting professionally).
It's not universal, but it is common. It depends a lot on the nature of the relations between the PI's and their groups. I tend to ask students/post-docs to keep me in CC of the first few emails to our collaborator after they arrive/we start the project, then after sometime I tell them to drop me when it's evident things are going well.
In your situation, the safest course of action is to keep people on the CC list. You didn't CC them, but someone did, so don't drop them unless they ask for it. Also, ask your own boss how he likes to do things.
“I imagine that as a group leader you get flooded with emails, so why would you request more?” — Whatever you do, you will be flooded with emails. You don't read them all, but you may skim those for tone/content, keeping an eye on things.