Another totally irrelevant crop vs ff test. I thought these had died long ago with the 7D vs 5D II debate.
You are making the assumption that in camera firmware is the same and that sharpness set to 0 on one camera is the same as 0 on another.
It doesn't account for camera shake in the real world which can quickly negate any benefit of the smaller pixels.
Neither is the real world a flat wall and evenly lit.
These kind of tests oversell what you can really expect. If you want to find out what the real difference is set up the test in a real world situation with real world conditions. Outside with 3 dimensional subjects in real conditions that you normally shoot, processed to the best of your abilities.
Enough bashing your test, I would say so far I have found from testing mine that the 7D vs 5D II debate has narrowed since the old tests a few years back. The crop factor is closer and worth consideration but so far I am leaning toward the negatives of a crop sensor being to great. The biggest negative is the loss of light and high ISO noise, this translates in the real world to a slower shutter speed. Slower shutter speed translates in to additional camera shake.
Your comments are the totally irrelevant contribution.
1. They were taken in RAW with no in-camera sharpening or noise reduction, and your comments about assumptions about camera firmware are out of place.
2. The focal lengths used with crop are 1.4 x shorter than on the FF so the effects of shake are the same on both.
I have also done similar tests with real birds in real situations with the same results. As someone else wrote, you test under controlled conditions and use what you have learned from those in the field. If you ignore such tests, then the loss is yours.