Many thanks to all who posted with thoughts - as I watched the responses roll in, my initial thought was "boy, if you wanted to demonstrate the shortcomings of your photographic knowledge to the world, this was the way to do it..."
Anyhow, I think the primary note I took away from my initial post was that my sample photographs were incomplete in their construction - that I'd failed to control for focus, and hadn't considered the distinction between focusing to infinity, and something closer. I also had never encountered the term "focus breathing"...
Anyhow, this morning I went out and made a new set of awe-inspiring images, trying to make sure I incorporated some of the advice. I've attached them below:
- Image_4818.jpg is the 24-70 F4 IS, shot at at manual, 70mm, F4, 1/320th, ISO 200
- Image_4819.jpg is the 70-200 F2.8 mk II IS, shot at manual, 70mm, F4, 1/320th, ISO 200
Both images were made with the camera mounted to the tripod via a baseplate on the tripod mount of the camera (I didn't use the tripod ring of the 70-200, in other words). The tripod and camera didn't move between images, so the focal plane (image sensor) was stationary. Obviously, because of the difference in lens size, the front element of the 70-200 is three inches "closer" to the scene than the 24-70.
The center focus point of the camera was placed, not on the green park sign, but on the trees/house immediately to the left of the green park sign. I visually confirmed that the lenses appeared to be focused all the way to the infinity mark.
The images are uncropped and unretouched, just downsized from the RAW file.
To me, the results continue to be interesting for what they show, even if I can't really explain fully why. To my eye, the 70-200 is slightly narrower (more "tele") in its "70mm" position than the 24-70. I now can't see the more significant depth-of-field distinction that the original (less-carefully-constructed) images of the Christmas lights seemed to show. At 100% in Lightroom, I can see some differences in image quality, but that wasn't what I was testing originally. And most puzzling to me, there seems to be about a 1/3 stop of difference in light gathering - on the second image, the camera reported the scene to be +1/3 EV, while the first image was +0. There are literally only a few seconds between the two images (long enough to do the lens swap), and it's a grey overcast morning - I didn't observe any shift in the ambient light conditions that would have explained the difference. (How's that for scientific? No? Not scientific? Well, fine! So it's anecdotal. Whatever.)
Anyhow, thanks again for everyone who offered knowledge - I appreciate it very much. Now I just have to decide whether I'm keeping the 24-70, or returning it for the 2.8 non-IS version...