Oh I like sciencey threads. This old physics relation shows what many of us already know from experience or intuition:
Resistance (R) is directly proportional to the length (L) of the tube and viscosity of the fluid (a constant in this case obviously). R is inversely proportional to the radius to the 4th power.
So as length increases, resistance increases (we know this already). What isn't as obvious is that as radius of the tube increases, resistance decreases
dramatically.
Of course we need more complicated math (computer models) to see how turbulence is created by bends in the tube and how that affects resistance. The flow through the filters could even be explained by the above relation in a sense, but it would require calculus and lots of data we can't get.
Basically, the large area of the K&N filters used on the Gen3 overcomes the negative effect of having longer tubes (I think their radius is a little larger than stock as well? correct me if I'm wrong). The lower IAT due to the filter location is not what gives the noticeable power increase. The difference in temperature between the stock draw location and the Gen3 location is too small to affect air density enough. Perhaps the computer is responding to the lower temperature readings and making adjustments that translate to more power?