If you can point me to that thread, I would love to read it over. I don't remember it very well (I might have missed it).
However, I don't think their discussion quite covers this situation. Noah was probably talking about comparing the contrast ratios of regular and gray screens. Don's test was not comparing regular and gray screens though, it was comparing a small sample of a regular screen and a gray screen. The situation is a little different, instead of measuring values off a full size regular screen, the values now come off a very small sample. If you have time please read my post in this thread http://www.avsforum.com/ubb/Forum9/HTML/000583-3.html
after Don posted his results. It explains why I think that Don's use of a small sample instead of a full size regular screen makes all the difference. The short of it is that with a small sample the ambient light would be the same for both the gray screen and small sample measurements.
I'll attempt to elaborate here:
Mr. Wiggles' post in the same thread shows how a gray screen can increase ANSI contrast ratios. This is because there is less ambient light bouncing around the room, which makes sense right? Less ambient light to pollute the black level. The same decrease in ambient can be achieved if you make everything in the room darker instead of using a gray screen, this likewise decreases ambient light. In fact, if you made your room very light absorbant you could increase your ANSI contrast without a gray screen. I believe the consensus that Noah's thread probably reached was that, while ANSI contrast is enhanced by a gray screen, actual contrast in the abscence of ambient light is not. Is that right? The increase in ANSI contrast comes from the difference in ambient light from a regular screen and a gray screen.
However, if you finagled it so that the regular and gray screen both had the same amount of ambient light when you tested them, you would find that their ANSI contrasts would be the same. The is what Don did when he used a small sample in his test. What cannot be explained by ambient light is why the ANSI contrast ratio was different for the small sample and the GrayHawk screen when the ambient level of light was the same for both measurements.
Not the best explanation, I know.