The central problem with all these tests is that, in dramatic contrast to a block of gelatine, the human body offers an almost infinite number of variables that promote or inhibit penetration.

A few years ago I and a few friends got into a discussion about this at a social event with a guy who was an MD and pathologist. He got out his laptop and showed us some 3D modeling. At a certain direction and angle, a bullet could zip through center mass without touching any bone and exit with comparatively little loss of energy ("overpenetrate"). Move the point of impact or even just the angle of impact an inch sideways, and suddenly there are multiple ribs in the path that increase resistance several-fold. Take a sideways shot that first goes through a limb (like the infamous Miami 9mm round), and suddenly it does not penetrate far enough.

Simply put, a round with sufficient penetration for all anticipated circumstances will inevitably overpenetrate in some situations. Any round designed to never overpenetrate will not go deep enough in some situations.

Anyone who thinks he can derive definite answers about better or worse, or safer and less safe, defense rounds by comparing whether they penetrate an inch more or less in a controlled medium is seeking false reassurance.

As already said, accuracy is a much more controllable factor than bullet behavior. It's probably more beneficial to focus on that.