Make sure any tank crossover valve is Closed. When crossovers are open, the fuel will tend to follow the path of greater suction and least resistance. My winter storage yard sometimes opens these valves up, I think just to encourage such mysteries as yours.
How are you judging GPH rates, by FloScan or other inline flowmeter which may need re-calibration? or are the flow rate readouts sourced directly from the ECC/ECM to Smartcraft or similar device?
If using inline flow sensors, are the fuel systems return or returnless type?
Have you seen any sheen in the exhaust wake at the dock that might indicate a leaky injector or other combustion issue?
Is the 5% GPH differential confirmed at the pump at each full fill-up? - Port engine always take ~ 5% more gallons than stbd., topped off?
Do you most always use an active synchronizer when not maneuvering in close quarters? Otherwise have you checked tach readings with a portable tach tester or marine scan device? ( ...The latter of which would also show ECC algorithm-derived GPH rate).
Is there any port lean underway?
Just to confirm & clarify , I think you said this is a relatively recent development but it was evident before the prop work was done and has not changed since those props were reinstalled. That seems a bit odd, since the props' pitches were apparently finagled in opposite directions, so one might expect to see ...some... change in same-rpm/same-sea-conditions GPH rate differential. Maybe just not very noticeable.
