Mark,I don't think that's really efficiency gain.
The lesser the flow, the higher the temperature drop for a fixed duty or work performed by the chiller. In other words,yes you are getting cooler water but less of it,which will the same amount of time to cool your tank.
Q removed = F*cp*deltaT so for fixed Q,you can play with F and alter T
That's also why the ranges are so broad,because anyone can find hissweet spot,which is usually the max flow you can flow through the chilller if it's inline.
By the way,I called the electric company(Reliant) and said they can do a calibration test for free if it has not been done in 4 years.great I say,otherwise I'll be happy to pay the $6 anyway
It turns out that their records indicate this meter has not been tested since 1964 8O
I'm hoping for some seriuos deviation here!