I responded too quickly to the last post and I should never claim 100% certainty in anything, but let me explain further.
The tanks are underneath the aft bunks just forward of the bulkhead separating the engine compartments. There is a fairly large diameter fuel line connecting the two tanks, entering at the top of the tank. On the port side there is a transfer pump mounted low in the engine compartment below the starter battery. There is a shutoff valve at the top of each tank right were the transfer line exits the tank. Moorings leaves these valves open, understandably, because they DO expect charterers to make use of the transfer pump, because charterers apparently run the genset a ridiculous amount. They do not want charterers to have to open and close valves under the bunk in addition to just holding the fuel transfer switch at the helm as the briefers instruct.

I did not open the tank inspection ports, but I can safely assume that the port (source) side must have a pickup tube extending to the bottom of the tank, or the pump would just suck air. I am not entirely certain if there is also a tube on the starboard (destination) side or if fuel pours into the tank from the top where the line is mounted. In the former case, a siphon may be created. I suspect this is the case and it may contribute to the "problem".

I missed my chance for the moment of verifying that experimentally by observing/listening to the starboard tank while the transfer pump is running.

I have also not yet verified that the Jabsco model transfer pump has an integrated check valve, but I believe that would be common in a low-pressure fuel pump to prevent backflow.

The unintended transfer does not seem dependent on the boat heeling one way or the other, only motion that causes vigorous sloshing of the fuel.

This all has been happening as long as I have owned the boat (3 years). On the rough crossing the other day, somewhat over 10 gallons of diesel in the space of about 2 hours moved from the port to the starboard tank.

I strongly believe that the sloshing of fuel creates pulses of pressure differential at the transfer line termini that, in combination with the check valve, cause the "spontaneous" fuel transfer bit by bit.

There are two further possibilities I can see, but they appear remote.

1) Someone at the helm (me) is inadvertently pressing the transfer switch, maybe by putting their feet up on the panel, and doing it long enough to move a significant amount of fuel, AND does not hear the pump running. AND I am only doing this on rough passages.

2) An electrical fault is causing the pump to be activated and the noise is going unnoticed and it only happens on rough passages.

That's this early mornings TLDR :-)