Ok, I can see what the thinking was there, re multiple players. We have always contacted both authors, unless they themselves set up a 'principal contact' as you say.
The thing is, this present set-up leaves it out of our control for communications to get to all authors - now, for important messages like editor decisions, we need to wait for the principal contact. And, hope that the principal contact is on the ball, and/or actually forwarding the correct info, etc.
And, like I noted, using the general 'author contact' seems fine until you see that it upsets the point of having the 'editor/author' correspondence tracking.
Am I right in suspecting the model for 'principal contact' you were thinking of was a natural sciences one, with many authors, some only tangetially involved in the final process? We're a humanities/social sciences journal, and 2 co-authors is the most we've had, both equally involved in the process.
Perhaps something in the journal management in 2.2 could be added - like a toggle to select a 'principal contact only' vs. 'contact all authors' model, per our own policies & disciplines.