... and at all higher ISO speed settings, the difference goes away.
Indeed it does. However, I happen to find ISO 200 one of the more frequently used settings. If most of my shots were at ISOs of 400 and above, I wouldn't be so concerned.
I recall when I switched from using the Canon 5D to the Nikon D700, I found that the higher base ISO of the D700 a tremendous benefit. At its base ISO of 200, the D700 had a full stop better DR than the Canon 5D had at ISO 100, and furthermore the D700 at ISO 200 was not worse in any other parameter, such as color sensitivity, and still marginally better with regard to SNR at 18%, compared with the 5D at ISO 100.
Using a shutter speed that's sometimes unnecessarily fast is not usually a problem and will tend to result, overall, in fewer rejects due to camera shake and/or unwanted subject movement.
A single measurement like that which falls outside the trend of all other measurements (and what we might expect based on the evidence that the two cameras use essentially the same sensor) and comes from measurements on a single sample of each camera, should be treated with skepticism. It might be real (weird Sony firmware?), but it is also quite likely to be due to measurement error or sample variation.
Quite so, which is why I asked the question, "Has anyone noticed that the DR of the A7r at ISO 200 is not as good as that of the D800E?".
If one suspects that a DXO reading might be a glitch or an error, it's advisable to test the situation for oneself, if one is able to. That's what I do. For most of us, it's the practical significance of those DXOmark readings that count.
One should also bear in mind that there are previous examples of Nikon and Sony using the same sensor in their respective models, yet the test results have been significantly different in some respects. So we know that there can be other features in a camera, whether software, firmware or hardware, that can affect the final result from the bare-bones sensor.
Also, if this 0.7 notch worries anyone, it can be avoided by setting the A7R to ISO 100 while still choosing exposure levels as if one were at ISO 200. (Ray, you are aware of this approach, are you not?)
Indeed you can. If the DXO test result is correct, then a 1 stop underexposure at ISO 100 should result in the same DR as the same exposure used at an ISO 200 setting, but that DR from the A7r will still be 0.71 EV worse
than the DR that the D800E produces in exactly
the same circumstances at an ISO 200 setting.