That's a fantastic graph there -- I'm going to save that link for safe keeping.
For whatever it's worth (not much), my own data looks more like a shotgun blasted it a couple of times:
The reason mine is so goofy is probably because it hasn't eliminated all the other variables such as yeast strain, mash time, amount of crystal malts, etc. This is just a simple plot of every recipe I have made over like the past 10 years, regardless of these variables. I suppose sometime I might want to whittle it down to more same/similar variables. But anyway, there you have it.
EDIT: Looking at the data a little more.... I can see that over the range of 144 to 162 F, the Kai graph would show a decrease in attenuation of about 12%. My own data for that same range of 144-162 F shows a decrease of 9%. Which is more accurate? Probably Kai. But suffice it to say that there does seem to be an effect, which seems to really kick in right at around 155-156 F. Below that mash temperature, it's pretty much a straight horizontal line, no matter whose graph you look at. Interesting...........