I get the notion that he's trying to control experimental variables
Reality is that when it comes to drilling, appropriate speed, feed rate, and optimized drill point geometry are critical.
For example, there are at least a dozen different drill point geometries (118° point, 130° point, 140° point, various split points and web thinning techniques, brad point, flat bottom, variations on margin thickness, lip clearance angle, chisel thickness, etc) that I could sharpen by hand, and each geometry would result in a different ideal speed and feed for each material. And we haven't even discussed coolants, cutting oils or optimal peck drilling cycles.
So when he standardized these values for the test, he eliminated the optimal conditions for each tool. The result is a set of standardized inputs that ensure substandard results.
Optimal drilling results require carefully selected and tightly controlled inputs that are different for each tool. This test was the opposite of that.
I can appreciate the efforts he made, but the results are just junk.