good to see it does, but what sort of difference in MPG could this make? is it significant or minimal?
decoupling the spinning bits with mass? IIRC about 10% is a reasonably middle of the road estimate. It really depends on various ratios, mass of the bits, the rolling resistance of the tires, and how much of the mass the rear tires are spinning in a purely parasitic setup.
That's if you are comparing traditional AWD vs something that decouples. If you are talking "let the computer sort it out" vs. "I pressed the just leave it unlocked button", probably pretty small to negligible.
As a point of reference, back when people were trying to figure out how to make 1G and 2G eclipses and talons go fast, pulling all the driveline from the transfer case back resulted in about 25% higher dyno numbers. Of course with that you weren't dragging the rear wheels, taking losses from bearing friction and tire deformation, etc. Manufacturer claimed differences between FWD and AWD were 20%, reality was less than that, and you weren't taking into account the drag from gears in the transfer case, rear differential, and cost of spinning the prop shaft. Some of the geekier engineering geeks tried to come up with an answer, and that boiled down to less than 20% but more than 5%, which was pretty much a "well duh, we knew that" conclusion.