From what I can tell, the important distinction is that all actors (and their messages) are described alongside each other, instead of being described separatel

From what I can tell, the important distinction is that all actors (and their messages) are described alongside each other, instead of being described separately. There are many implementations of the actor model, but most of them are the 'static-location architectures' that TFA talks about.