That is proof positive that it's not simply because it's in 4th gear. The initial variations had the fuel rail jumping from 5K psi to 9K - that's an amazing variation - and in 2HI it's probably not scratching the 7500-8500psi mark in variation.
(For the avid reader accustomed to petrol injection fuel pressures of 80 to 200psi, the common rail diesels in these cars have extraordinary fuel pumps that can drive the rail pressure to 30,000psi which is nearly twice the pressure at the bottom of the Mariana Trench, the deepest part of the ocean on the planet).
Much of that variation is going to be the ECM's variations to provide test data for the oxygen sensors in the exhaust.
It looks like in 4LO that the variations are exaggerated. Why? Why in 4th gear, and why only around 2,000rpm?
The ECM knows if you're in gear or not (it has to, because it inhibits the choice of 4LO unless you're in neutral). I don't think there's a variation in the gearbox between 2HI/4HI/4LO - that's all inside the transfer case. Nothing inside the gearbox should change.
And so we come back to a logic problem. Are they using a CAN signal that is incorrectly identifying itself, or is the signal coming from the transfer case sensors cross-linked into the HO2S in the exhaust, or the turbo actuator (vacuum)?
It'd be wonderful if you could manually control the turbocharger vanes, and disconnect the vacuum lines and still have some boost. That would at least discount that.
I wonder if FB could have a squiz at the internal database and see if there are any issues relating to this?