I just thought I would add here that eye-balling graphs are only good for rough guesses. In scientific research you use mathematical tools that show correlational relationships. There are well-defined models and standards this way that give you confidence levels. My point in making the graph at the top of the post was not to try to prove anything but to just show a bit of my reasoning behind my predictions.
Oh yes, we did use the average publisher figures instead of the peak publishers.
Skipper