Like resistors wired in parallel, coolers plumbed in parallel reduce the flow resistance to a figure lower than either cooler.

Just as adding more tubes to a water radiator reduces the resistance to water flow.

 

Like making a fixed length of tubing bigger in diameter as you add coolers. It flows better the bigger it gets.

 

As in resistors wired in series, coolers plumbed in series, just add total resistance to flow of each cooler.

 

Like making a fixed diameter tube longer and longer as you add coolers. Eventually resistance to flow will result in no flow at all.

 

In addition: The hotter of two parallel coolers will flow more fluid, as the fluid becomes thinner. So, in order to balance the flow between one leg of a parallel system that has one short fat leg, with another leg that is long and thin will be one interesting process. 

Perhaps restrictors or manually tuned valves will be involved

 

Plus, the rotary is cold blooded. The big bearings stress the oil film to no great extent, and the major source of oil temperature is rotor cooling. At low speeds and idle, the check balls in the crank don't even open to allow cooling oil to spray into the engine. Those balls operate as a function of RPM not temperature. 

 

On the other hand, you will get a water temperature increase within one minute of startup. A water based coolant has very low viscosity and flow to a distant heat exchanger will be immediate. Like the rear heater in my school bus.

 

There is so little difference in weights that there is little disadvantage of one over the other

(Oil vice Water).

 

Oil changes with the "long" system will never get a complete drain cycle.

 

Lynn E. Hanover