We seem to be having a communication problem, so apologies ( I mean in this situation you are discussing) - at 5:00am with no coffee, my comm skills are obviously suffering a bit. So I shall try again. Here, one is considering a 28-85mm lens, you personally have a 24-70.
TBH, that range is not one I work with - which is not surprising, we all use the tools we need for our particular genres. I do have the EF 17-40L f/4 and have actually shot with it on my R-series bodies and it worked well enough for my purposes. However, I got the 14-35L and am very happy with its performance. I am otherwise happy with the RF 24-105L f/4. Still, I think this new 28-85 f/2.8 will fill a niche. Indeed, you are right, it does remind me of the FoV one would have got with the EF-S 17-55 f/2.8 for crop sensor DSLRs, but since moving to the R-series bodies I have stayed purely with FF cameras.
Although some folks are not happy with the use of computational photography being applied to enhance and correct the image generated by the optics, I see this as dedicated camera lenses finally catching up to the cell phones, in that they have been using computational photography for years, and have leveraged that to pretty much massacre the point and shoot market for the masses. I don't really see it as any different from the rather significant firmware and software that is in cameras bodies to focus, track and stabilize subjects and cameras. If it allows the makers to deliver smaller, lighter and cheaper optics than without such apps, then I am all for it. Another advantage is that the firmware can be updated so, the algorithms can be updated to fix bugs or tweak improvements. I have seen that with such lenses as the Sigma 150-600c and 60-600s lenses and is appearing in many brands to varying degrees.
