Whilst I see what you're saying this time the difference really is hard to tell (so hard DF had to be within a foot of the TV IIRC). Certainly I think as Scorpio (sounds like) it would scrape into 4K rather than do it with ease I would have thought it's got to be a better to go checkerboard and add some nice effects.
Base Scorpio specs land it squarely to be able to take theoretically _any_ XBO game and make it run natively at 4K. The implications there are pretty big but are often ignored because everyone wants to use that power to make graphics better and resolution no longer matters, and not to anyone's fault, but MS marketed it that way.
The biggest advantage here is that any XBO game can become 4K with relatively small effort;
looking back at the existing 3 year catalog of XBO games, any developer can go back and make some tweaks to the graphical buffer to 4K remove esram calls and you've got a 4K game. This would also apply to any developer who cannot leverage checkerboard rendering techniques or does not have the time schedule to meet their deadlines to make a vastly optimized 4K reconstruction renderer.
For developers who want to do more, or are working on future titles, the expectation is that they will leverage reconstruction and use the remaining power on other things.
MS does not care if it's 4K*, they only care if it's vastly superior to its competition in which reconstruction would enable.
MS walks away with some big wins here in this department, Scorpio will be the king of graphics at least until next generation arrives and it can (I think) launch with an enormous existing catalog of games patched to be 4K. Do not be surprised if this E3 there is a massive catalog of 4K Scorpio games ready.
4Pro requires a bit more work to make this happen, it ultimately does not have the hardware to run a game at 4K natively, at least it has no chance at running a PS4 game at 4K without reconstruction;
tldr; I would expect the result to see more 4K games (through w/e rendering means) on Scorpio than on 4Pro.