If PMs work the same way as “universal inline whispers” would, why do you need those whispers. Obviously, they don’t work the same way and the difference is what makes a difference here (see party analogies above).
This is actually a very nice example for illustrating my concerns: nobody has said that this would be a conversation killer. If your interpretation “People seem to think, this would be a conversation killer” is posted publicly, anyone in the conversation can adress it as a misunderstanding (or agree with it, for that matter). Even if no one actively relates to that interpretation, it is still a valuable contribution to the discussion because if makes transparent what you think of the conversation so far. And if Dann later joins the conversation in your support, it is transparent that he does so upon your request.
Consider, by contrast, the situation where that whisper indeed were hidden. Chances are that you and Dann would continue your whisper conversation between the lines based on the assumption that “People seem to think, this would be a conversation killer”. By consequence, the two of you might be developing your own parallel discourse that nobody knows anything about and that is the exact opposite of dialogue.
Dialogue requires that everyone involved makes their thinking transparent so that everyone else can relate to it and try to understand it.