No, we don't have actual comparison data: that would be lenses tested on the same sensor. We have different data that you are using to extrapolate a result.
I am not trying or desiring to defend these Sony lenses, but I do get frustrated when I see these wild accusations that one product is "far inferior" to another. We're going to see the same hyperbole when Sony/Nikon announce their 50 MP cameras. The poor 5DS R will be regarded as "far inferior". And it won't be true.
I had a D810 with 14-24 and an a7R with the FE 16-35mm. The 14-24mm enjoys a great reputation, but when I tested the two lenses on their native 36 MP sensors I struggled to find much difference in real-world scene tests, and I did a lot of pixel-peeping. I became so comfortable with the FE 16-35mm performance that I sold the D810 and 14-24 (just preferred the smaller size). I wouldn't be surprised if the FE 16-35mm was not able to hold an equal DXO score to the Canon version if tested on identical sensor because the FE 16-35mm drops off a good deal at 35mm. But, I don't care: it's a 16-28mm lens for me. I have two 35mm primes that I will use before using the 16-35mm. These Mpix scores for zooms are likely averaged over the whole range, so they don't give any sort of detailed picture of the lens' performance.
I own/use two of the top two DXO Mpix scored lenses (Zeiss 1.4/55mm ZE, and 2/135mm ZE), so I know what amazing lens performance looks like, and yet I also enjoy my Sony FE 24-70mm, which is supposed to be a turd of a lens. I get sharp images from that lens on my a7R somehow.
Well said! This matches pretty well what my experience is