OK - if you guys -really- care about this stuff, go get "Tune To Win" and read Chapter 9. the edition i have is a bit dated and technology marches on - FluiDyne is now the Hot Setup for coolers. technology changes, but heat and air work pretty much the same way they did 30 years ago...
he actually doesn't elaborate much on the parallel setup, he just says do it - for efficiency.
since we're looking at total heat rejection, Delta-T is significant. assuming we're dealing with multiple coolers rather than one big one for packaging purposes, for a given Delta-T, two coolers in parallel will be smaller and lighter than if they're plumbed in series. put it another way - they'll reject more heat and "cool better".
that's my story and i'm sticking with it.
i wouldn't mind participating in a controlled test, but it has been done before...
now - that said, i will be plumbing the supplemental automatic transmission cooler in the Subaru (light-duty tow car) in series with the existing cooler. here's why: i will be adding a freestream air/oil heat exchanger upstream of the factory cooler that is actually part of the water radiator. by cooling the transmission oil when it is at its very hottest with air, i will be removing some of the heat rejection load from the engine's air/water heat exchanger ("radiator") under the conditions that it is also working at its hardest. in cold temperatures, if the trans fluid is cooled below its optimum by the airstream (which i will be able to block off), its trip through the engine radiator should actually warm it up a bit, since it is downstream.