Yes, I have noticed it. The difference is less than 2 DeltaE 2000, the difference is normal.
In recent Eizo CG and NEC SpectraView (27”/30”) you get the same exact casts (is green/magenta) when you have a white or light gray background. I think that this is normal, if you can measure the DeltaE between center and corners and you find that is less than 2 (usually 1,5) than it is normal. Anyway it varies from screen to screen and it could depend on calibration settings and on the profile that you are using (check also the cable).
One method to obtain measurement is imply to do several validations in different areas of the screen: the first one on the center of the screen, than a second one on the right side and a third one on the left side.
I am making some tests and I think that I will publish them in the next days, to show what I call “unwanted color gradients” and “unwanted light gradients” (check also luminance, on the Cintiq 24 HD one side is usually 8-12% brighter than the other).