I wouldn't take the two reviews cited above as a reference for judging the performance of these lenses. There isn't a lot of information about the methodology used for sharpness measurements, so it could well be that the results depend on image processing. It might also well be that vignetting affects the measurement. It is surprising that center sharpness reaches its maximum at f/4, while edge sharpness reaches its maximum at f/11 (!) for the 15-35 at 15 mm. Most recent Canon designs (or maybe even all of them) didn't need that much stopping down to reach peak performance even in the corners, and it would be very surprising if that changed with the RF lenses. Something is a bit off in my opinion.
Also, the TDP results for the 15-35 do not look disappointing to me. The comparison with the EF lenses tested on a 5 DSR is difficult, as confirmed by Bryan Carnathan in his comments to the review. He also suggests comparing it to the RF 85 tested on the EOS R. The comparison is then between a UWA zoom lens and a 85 prime (possibly the best one available), and that doesn't look too bad.
Another aspect is distortion: The 15-35 has significantly lower distortion than all EF 16-35 designs, and it has also lower distortion than the EF 24-70 in the overlapping focal range. Software algorithms are getting better and better, but I'd argue that having slightly less edge sharpness but significantly lower distortion might still lead to better detail in the corners than with a lens requiring a significant amount of distortion correction in post.
There is a video "review" of the 15-35 here:
It's not a scientific test, but it compares the RF and EF lenses in real-world situations and there are some interesting aspects: The 15-35 seems to be only slightly less sharp at 15 mm than the EF 16-35 III at 16 mm, but at 15 mm there will be "fewer pixels available for the same amount of information", meaning that with identical sharpness a 15 mm lens would look worse than a 16 mm one on the same sensor, if the 15 mm image is cropped such that it covers the same fov. So, it's possible that the two lenses would have the same sharpness at 16 mm.
What also seems to be good is coma performance. There's just one image, so again not a scientific test, but the corner performance of the RF lens at 15 mm looks at least as good as the EF 16-35 at 16 mm if not better. If that turns out to be true it would be great!
Vignetting is difficult to judge, but it doesn't seem to be worse than the EF lens. That one is rather bad concerning vignetting though, so this might be a less than optimal result. But then, the 15-35 lens allows the use of front filters and has a not so critically exposes front element while other <16 mm f/2.8 UWA zooms with less vignetting wouldn't.