I actually thought the OEM intercooler was pretty good, even in a highly modified 335d. Here's some data that shows comparisons of BT recorded data of the IAT (post intercooler) for pulls done with the OEM intercooler and with an initial setup of water injection, and then the final setup with water injection.
The OEM setup on a pull on the road (not a dyno, which will have much less air passing over the air to air intercooler than in real world applications) doesn't leave a lot on the table for a highly modified car. And the water injection was able to bring the IAT's below ambient for a time (something no intercooler can do).
Regarding the dyno data provided, since it's on different days, with different fuel, with different DPF loading conditions, is the few hp differences statistically significant? For example, here's a test of the same car, same day, same fuel on several dyno's ... all different results.
Regarding the mpg impacts ... my car at 70 mph on level ground isn't using any boost. The flow rate of air moving through the intercooler is pretty low in this state, and the IAT's are basically ambient levels, so I'm not sure what laws of physics could possibly be involved to explain the radical increase in thermodynamic efficiency being described.
Not trying to ruffel any feathers, just trying to give some insight and quantifiable data that was collected in the past.
Peace fellow diesel lovers :-)