>Look at the first three images, all with the same camera >settings in low light. Which is the better image and why? Do >those images support your theory or contradict your theory?
They don't address the "theory" at all because the subjects are not in motion. When subjects are in motion, you have to have a shutter speed fast enough to stop the motion. That fact isn't altered by any new technology, explicitly not by VR since VR does nothing to stop subject motion, only camera motion.
It's certainly true that camera technology allows shooting at ever-higher ISOs, but that only means that we will be trying to shoot in ever-darker venues with moving subjects, so capturing as much light as possible will still be needed indefinitely.
Note that I'm not even bringing up the fact that in low light, an f/2.8 lens delivers more light to the viewfinder and autofocus system, making for more reliable low-light focusing, nor the fact that the DOF is narrower on an f/2.8 lens, producing a photographic effect unavailable with the f/4 lens under the same conditions. (Well, OK, I guess I am bringing it up.)
I'm delighted to see the 70-200/4 lens appear and be as excellent as it is by all reports. I'm sure for lots of people it is a better choice than the 70-200/2.8. But there are times when an extra stop matters.