I took an even closer look and have changed my theory of what is occurring. My earlier theory was based on the timings reported by toc as echoed on the command window. In that experiment, I just looked at the timings in the command window & estimated that inter-mouse event intervals were approximately 0.03 seconds (with some variation of course). The same experiment in version 2013 produced inter-mouse event intervals were approximately 0.015 seconds.
BUT.... this experiment was flawed in that: 1) It relied on just visual inspection AND (more importantly) 2) toc was echoing text to the command window.
So... although there WAS a huge difference in timings, this difference may not be the result of delays from mouse movements to the associated mouse movement function.
So I changed my code & stopped toc from echoing to the command window and just stored the timings in a variable which I later wrote out to a csv file. I ran the code in both 2013 & 2014 (moving the mouse as fast as I could) and found essentially NO difference!
So.... the problem lies.... elsewhere. I guess I'm back to my original theory that version 2014 is slower in interacting with graphics (or even the Matlab GUI) than version 2013. I will continue experimenting.
Dan