Defining steepness is not that exact. I think what people are interested in is the steepness of the steepest "sustained" pitch, not the average steepness (which in many cases includes a run-out). The definitiion if "sustained" is what provides enough wiggle-room for flame wars. Obviously the back side of one mogul is not long enough. But what is?
You can get an approximate steepness from topographic maps, but the fine-scale detail won't be there. And snow accumulation changes things sometimes (at the limit, think cornices).
In my opinion, the three valid, accurate ways to measure are:
a) stop and meaure it with an inclinometer or using a ski-pole protractor. But even then, how many different spots do you measure? At the one that looks steepest?
b) photos with true verticals (e.g. trees) for comparison. This relys on a competent and honest photographer, since it is easy to compose deceptive photos. But I have taken such pictures for my own curiosity where I was careful to get square to the surface and the trees and I think they were pretty accurate.
c) analyze GPS logs. As always with GPS, you need to exclude wild points and do some averaging over a span of several points to get accurate results. (That is, the speed or slope computed from adjacent log points is not reliable by itself). The averaging is a good thing, though, since it goes back to the "sustained" thing.
For my own preferences, the "sustained" window is pretty short -- basically long enough to feel it, maybe two or three turns. Others may wish to follow a different philosophy.
(If you think about it, you'll see that a longer window will give a smaller slope.)
My impression of the skistats site was that it was a nice starting point and general overview, but not an authoritative reference.