Maybe I'm not following your argument, KVH. The pistons do pull the clutch away from the annulus; but until the clutch locks against the brake ring, it will just spin uselessly and there will be no power transmitted through the gears. The sprag (roller) clutch inside the annulus normally covers this transition, since it keeps the input shaft from turning faster than the output shaft (so effectively the OD stays in direct drive until the OD clutch locks to the brake ring).
However, if the pressure in the pistons is too low, the clutch doesn't fully lock to the brake ring and can slip back to direct drive (through the sprag clutch) when extra torque is applied. I assume that's what tdskip is talking about when the "RPMs jump". Could be a mechanical problem, but most likely is low pressure IMO. I got some slippage at around 250 psi, but only at full throttle in 2nd gear.
Experiences will vary, but mine has been that for there to be a lot of sludge, the OD is coming apart internally and the sludge is actually very tiny bits of metal. I don't know why Jerry thinks it will accumulate in the wrong places, as it normally accumulates at the lowest point of the reservoir, which coincidentally gets removed every time you drain the oil. Or on the magnets, same story. The only tiny passage is that one in the valve stem, which is also easily cleaned.
Way I see it, if you pull it apart and it only needs gaskets, seals & thrust washers, then the job is cheap. But replacing hard parts gets expensive very quickly. Thus if you know it's coming apart, it's best to service it now rather than wait until later. Broken thrust washers let the gears start touching in bad ways, which will eventually tear them up even though the OD is still working. Somewhere, I've got a ruined annulus that came out of a "working" OD.
Randy, have you checked the hydraulic pressure? One cause for repeated clutch sticking can be too much pressure. And your DPO/M may have made the same mistake I did, by shimming up the pressure just a little bit too much, so it tests fine on the bench (at relatively low rpm), but the spring reaches coil bind before the relief port is fully uncovered (meaning that at higher rpm, the pressure rises until something breaks). Not the only possibility of course, but one of the first things I would check.
IMO, ATF would work fine as a flush. That's even what Herman van den Akker used when testing on his stand. But I would definitely avoid using full throttle in lower gears (or even doing a hard launch). The countershaft thrust washers in the main gearbox are the Achilles heel, and the lower gears are when the forces are highest on the countershaft. 4th gear should be fine, though, as it doesn't send power through the countershaft.