All that has been said is true. The science is this... the part of the IP that rotates and distributes fuel to the injection lines, does so by aligning a single port with one of the eight radiating lines. This alignment is gasketless, and relies on a tolerance that is extremely fine. When the plungers come together to squirt the fuel, the fuel must pass from the rotating part, into the stationary part, with the fuel passing through free space between the two parts. This space is so small that a fuel molecule is bigger than that space (Ok, that's an oversimplification but you get the idea), and has no where to go but into the opposing port and down the injection line. All of this, being done at over 1000 PSI. If the stationary part were to "grow" or say, get larger in diameter, then the fuel molecule has an alternate path to take, and it doesn't get to the injector. Heat soak, actually causes this. The stationary external head gets bigger, faster, then the internal rotating head, and the resultant space between the two parts causes the injection event to fail. Pouring the water on, shrinks it down and fixes the problem. But, you can see that if it shrinks too much, you have a serious problem. Once you get it started again, the circulating fuel evens out the temps in the pump, and all the parts grow or shrink at the same rate, preserving the tolerances. A worn pump's tolerances are too great, thus the heat soak becomes a problem, where as even on a good pump you have heat soak, but the space between the parts was smaller to begin with, so even after expansion the injection event suceeds. I can see how inserting really hot fuel into a pump could cause the same problem but from the inside out, expanding the rotating head faster than the stationary head could expand. Alternative fuel viscosity can be a more significant problem however, since if the plungers can't squeeze the fuel through the orifaces sized for diesel fuel, then your going to get a hydrolock situation which will also snap the shaft, and destroy the advance cam ring.
My understanding from here is that a new head in a DB-2 costs more than southeast is charging for their rebuilds. 
