Problem that you will find is as you start to increase tube dia. it will not be a linear amount of leaning out.
The change in flow is a function of the cross-sectional area, and therefore linear. The manner in which the air tumbles at low flow skews the way the MAF sensor reads, introducing uncertainty, and that changes for every permutation of intake tract. There are also some harmonics taking place that skew the way a MAF works, as well.
You see, your problem is that you ASSume the factory MAF curve is correct in the first place, which it is very much NOT. Fords apply half their fuel trims towards a MAF trim in order to learn the MAF, when you changed shit up it not only had the *wrong* MAF curve (despite whatever you were using to spoof the stock curve), but likely the learned MAF trim stacked on top of it.
It's kinda wierd you use a 3V example, 110% of those cars kits come with something to reflash the ECU that raise a bunch of max fuel pulsewidth limitations in the code. Kinda hard to do one right without directly editing the ECU for the changes made.