One thing that really strikes me looking at page 41, table 6 (4 blades honed on 1 micron diamond on the tormek machine) is what a tremendous variation there is between the blades. I don't think we can directly compare these to the chapter 4 blades, but they can be compared with each other. These blades also seem to have been analysed more rigorously than the ones in the rest of the paper.

# min/max avg
3-63 0.18/0.34 0.26
6-63 0.20/0.42 0.31
7-63 0.20/0.38 0.29
8-63 0.34/0.45 0.39

That's a range of 0.18 - 0.45 microns all from identical blades honed on identical 1 micron diamond abrasive. He also mentions doing the same thing with chrome oxide and 6 micron diamond and the edge widths increasing by 0.1 to 0.2 micron for both of them. But that's far less than the 37 micron variation we see here in this one situation. Which makes me wonder just how much any of these numbers really mean; I wonder if the sample size is too small.