I thought this initially, too. But after spending more time with it, and moving my head from side-to-side, certain spots do get
better at certain angles, but are never quite the same as other spots. The upper-right corner of my screen is always whiter than the bottom-left, and I don't mean relatively-so. If I look at the upper-right corner straight-on and up-close, and then move my head to look at the bottom-left corner straight-on and up-close (and from the same angle I was just looking at the upper-right corner from), the bottom-left corner does get brighter, but it is still not as bright or as white as the upper-left corner.
Also, I still contend that I'm not seeing grayness/shadowing. I'm seeing yellowing. The white-point is different on that part of the display. If it was just seeing a slight bit of shadowing, then I wouldn't be as...annoyed.
There are so many reports and pictures of this phenomenon on the rMBP, though, that it is really hard for me to believe that this isn't a universal engineering...well, "defect" is too strong a word, but perhaps "weakness?"...of this particular display technology or process that affects all displays coming off the assembly line. I previously owned two IBM ThinkPads with 15" IPS displays, and they both had a similar problem, but the "shadow" was on the right-middle side of the display instead. The
responses I got from others (as well as a replacement screen that looked
exactly the same, in one case) eventually led me to believe that all ThinkPad IPS screens actually looked this way, but that not everybody noticed or were as sensitive to the issue as I was, so I eventually got over it...mostly. If I had to be honest, it still bugged me every time it caught my eye and I became conscious of it again.
I want to just chalk this up to being a weird IPS characteristic that all panels have in common and which some people are simply more sensitive to than others: viewing angles don't cause colors to "wash out" and invert as they do on TN displays, but do cause whites to look different and the brightness to appear less intense. The only thing holding me back from believing this are the anecdotes from posters such as Davieis, who describes a similar problem on his own unit's panel, but claims to have seen other machines that didn't have the so-called problem.
-- Nathan