People in the GM 6.2-6.5L community report that running straight ATF has caused the head and rotor of DB-2 pumps to "hydrolock" momentarily and snap the shaft due to the increased viscosity of the ATF versus diesel fuel. These anecdotal reports are as valuable and useful as any that report a miraculous cure, if we are going to accept all anecdotal stories as factual and give them weight, then we must give them all weight.
In trying to think of why this ATF thing MIGHT work, I wonder if it's not the slug of viscous oil that is effecting a "cure." The ATF isn't in there long enough to do much from a cleaning standpoint (plus is isn't that great of a cleaner). According to my references, #2 diesel fuel has a viscosity of 1.3-4 cSt, depending on temperature. Warm, it would be at the low end and cold at the high. ATF is variable, but usually in the 20 grade area around 7 cSt at 210F (100C) or 36 cSt at 100F (40C). Given diesel fuel is going to be maybe 100F at the warmest, your pump is operating normally on fluid that is 1.3 cSt. So by putting ATF thru that is probably even COOLER than 100F, you are putting a fluid thru the pump that is 36 TIMES thicker than the pump is designed for. I don't know, but that seems way more that the viscosity compensator in the pump can deal with if it's expecting and designed to see fuel in the 1.3-4 cSt range. The thicker fluid can't be pumped as easily, so more stress is placed on the various components and the fluid either is pumped or something breaks in the process. If there are stuck parts in the pump causing a problem, and increased viscosity might force them to move.
So, Mel, what exactly happens to a pump when you put really thick stuff thru it? Have you ever put a pump on the stand and tested what happens from increasing the viscosity of the test oil? Who knows, maybe it's not good for your test stand but I'll be somebody has done it.