Mine was PMD, guys... after I changed it out, it never came back. Good thing, too; that surge-in-reverse thing is too hard on the shorts when you're in a parking lot full of Cadillacs and other expensive but fragile vehicles...
Others have reported that they cured theirs by changing ECMs. One guy by changing IP (armature shaft).
Another theory is, the idle system runs on a feedback loop; the computer senses the rpm and gives the command for more/less fuel to keep it where it wants it to be. If the IP starts to run out of fuel, the RPMs will drop and the computer should compensate by calling for more fuel. If the IP suddenly GETS more fuel, you got yourself a surge. If there's enough of a gap between the rpms dropping and the IP getting more fuel, you get a surge. If you push a steady-state system out of balance, it pushes back. If the initial push was hard (ie: almost stalls), the pushback will also be hard. That's why PMD failures cause surging at idle. They seem most obvious in reverse because of the line pressure and gearing... reverse is pretty direct-drive.
In other words, check the fuel filter, lift pump, OPS... look for something that interrupts the fuel flow.
And, I'm pretty sure I know somebody who will tell you it's air in the lines.
I could buy that if it happened on a regular basis. If it just happens sporadically, look for something more systemic than leaky lines.
Sorry I'm not more help. I even went and searched/read a pile of threads in the old place... nothing conclusive.
