Yes, it does seem a bit counter-intuitive that warmish water is more efficient than cold water. Here is a paraphrase from my system's technical manual that may help in understanding it.
The reason (that 90F sea water is optimum) is that in the refrigeration cycle, after the high pressure liquid refrigerant has passed through the expansion valve and gotten cold by converting to a gas, then passed through the cooling coil and taken on the heat of the refrigerated area and then gone back through the compressor, it is about 140F and only needs to be cooled to about 120F to condense back to liquid and repeat the cycle. Thus 80-90F cooling water is all that's needed. In colder water the liquid refrigerant is too cold and is stored in the condenser and the system operates with a much lower suction pressure as the pressure difference across the expansion valve is reduced. This will substantially decrease the capacity of the system and increase the running time.
That same manual offers corrective solutions for operating in cold open ocean water by using a gate valve or a water by-pass to restrict the flow of too cold water and get the drier and receiver temperature up to about 95F.
From just an everyday non-technical vantage point, consider that the marine market is absolutely insignificant to the global refrigeration market (no special refrigerants, etc. - there is no such thing as "marine refrigeration" outside of the marketing department). Nobody has cold water cooling - Eskimos don't need it! Household refrigerator cooling coils have ambient temps in the 70's or 80's or more (room temp). Apartments and buildings have water source cooling tower ambient temps in the 80's and 90's (or 100's here in Florida) and, therefore, cooling water temps are the same. Those systems are, in fact, at maximum efficiency because they are providing cooling at the condensing temperature of the refrigerant - 80 to 90 ish, depending on the refrigerant.
And, oh by the way, I just Googled Windy.com, clicked on the water temperature, and from 20 S to 20 N, including the entire Caribbean, the water is in the 80's. At my dock in Florida it's 92F. My A/C and refrigerator love it! As do all those cruisers who used to be there before Covid.
I would assert that no marine refrigeration system is designed to operate in extra-tropical ocean cold water temperatures - unless it has a gate valve or by-pass for cold water use. That pump life has been seen to be inversely correlated to ambient temperature is perfectly understandable. It's not the cold cooling water, it's the cool ambient temperature. The cold cooling water is actually increasing the running time and fighting against the cool ambient temperature that is decreasing the running time.
It would be fun to experiment with a system with a gate valve or by-pass valve in a cold place. Run the system with 55 degree water for a few days and log the run times. Then throttle down the valve to get 95 degree water for a few days and log that. Then check Oliver's fancy logging system and see what it shows. Dollars to donuts the warm water will see less pump usage.
Fun topic for a locked down Covid evening, thanks for initiating the discussion,