Since actual pig-based ballistic gelatin is expensive, heavy, bulky, and in other ways complicated to work with, he wanted a decent way to convert results obtained from shooting a row of common gallon milk jugs filled with water. Seems like better science than my rough estimates have come up with.
Final numbers are determined by inspecting the extent of damage caused to the final jug that the bullet comes to rest in and applying a modifier.
Each of the jugs penetrated before the final one gets a penetration value of 6".
If the final jug is penetrated on the front wall, with no damage to the back, add 3"
If the back wall is dented, add 6"
If the back wall is cracked, add 6.75"
If it dents the leading wall of the next jug, add 7.5"
If the leading wall of the next jug is broken, add 8"
Conversion factor: add the variables up and divide by 1.55 for handguns or 1.3 for rifles with velocities over 1800 fps. The result will be your rough equivalent gelatin penetration number.
Basically, if you're comparing to the FBI's "ideal" penetration range of 12-18" in gel, you start making minimum at damaging the boundary layers between jug #3 and #4, and you make the 18" maximum between entering and denting the back wall of #5.
I haven't wrung this out against a known-quantity duty round yet, but it does seem to jive with the 3-5 jugs most duty rounds seem to give.
Have fun kids! Stay dry!