I hate to be the one to rain on your parade... but, these "test" are meaningless because you have two tolerances to take into account (1) the wrench and (2) the nut - you might have selected a nut at it's minimum tolerance and a wrench at its maximum and that's going to change the contact area and if the situation was reversed (wrench minimum / nut maximum) the contact point would again look different.
While the results are interesting, they don't prove or disprove anything.
True.
He can try to minimize the number of variables. Even with the same nut, he'll have to keep it consistent and not rotate it and use the same contact face every time. No guarantee one face pair is the same as the other two face pairs on the same nut.
Better to use one nut, the same nut, the same face on the nut for the test with all the wrenches.
Or, have a quantity of X nuts, and for each wrench, test it on each of the x nuts.
You will also have the same problem with different faces of the wrench. To be consistent, the wrench have to be used with the same face each time.
In the end, it is just a very interesting observation.