I have a 54622D as well. It is a great scope, so well worth fixing up. I'm very impressed with the "mega zoom" on the delayed timebase, the ease of use and responsiveness. It is also really well built.
I wouldn't recommend using water on any parts other than the case, knobs etc. For the PCB, clean carefully with cotton buds and IPA. Still, I doubt this is causing the error.
Could it be possible that the 2 PSUs are not quite the same spec, or missing some extra supply line(s) on the 54622D over the 54621A? The former must have more logic devices in it. Howabout you focus on fixing the broken 54622D PSU rather than cleaning the main PCB?
There is a guy on eBay selling home-made logic leads for these. He has taken leads from an Agilent LA and then made a small PCB to slot into the edge connector on the 54622. They aren't cheap, but much cheaper than trying to get an original set.
BTW, does anyone have info on changing the battery in the 54622D? I did see some chat about it, but it wasn't clear to me if critical calibration settings would be lost if it were removed or not (like older equipment that stores setting in SRAM like HP3457A etc.).
If you do put it in the dishwasher please post pics