Sal, I have been toying with that idea myself. It would not be overly difficult to make a pressure sensor give the required information about cut depth vs applied load. I should stop putting it off and construct something.
The "various" materials" cut present tremendous variation. Magnified by total cuts, became less accurate.
Sal, as I told Bob Taylor :
http://www.bladeforums.com/ubb/Forum1/HTML/001940.html
Uncertainties or errors do not work this way. In fact they behave in the exact opposite manner. The more of a material you cut the more consistent your results become. If you did 100 cuts through cardboard and compared two identical knives the difference would be 10 times less than if you compared them after one cut. To understand this in detail just read over the section on sampling distributions in any introductory statistics text.
As an example, take the current 2000 members of bladeforums. Now if me and you each took 10 people at random from the total population of members and compared the average ages from our groups there would likely be a big difference. But as we took larger and larger samples, the mean ages of our groups would converge. If we each took 1000 people at random the means would be almost identical. This is the effect I am talking about.
We also control pressure to very close tolerances, difficult to do when using humans.
Even though a normal person cannot get really precise angle or pressure control of a knife, on any single cut, the mean effect of a lot of cuts will be very consistent. You can get whatever precision you want by just having him cut more stuff. This is also a direct replication of actual use which offers much more valuable information as to how the knife will behave for the ELUs.
Machines are all well and good, and very precise measurements can never hurt, however neither of them are necessary for scientific work nor do they strengthen an arguement by themselves. When doing any experimentation you always have to look at just what you are trying to figure out and make sure that it is logical to extrapolate from the usually very fixed lab conditions to the very dynamic real life situations.
For example the last time I did a prying test I was harshly critized by a friend. What I was doing was very carefully increasing the load until I surpassed the yeild strength of the knives. He pointed out, quite correctly, that while this was all well and good it was not overly meaningful in a general sense as I don't actually pry like that. So after I did my careful controlled testing I did some actual dynamic prying as the final step on all the knives that passed the first round.
-Cliff