From what I understand the Paxtons are much like a turbo (both centrifugal air pumps) in that if at 1000 RPM they produce 2 PSI, and you triple the RPMs to 3000, the boost will more than triple to say 7 or 8 psi. In short, the boost relative to RPMs is non-linear. That is why you would need some kind of popoff or bleed valve - to keep it under control at higher RPMs
Even if that is not the case and you can assume a linear relationship between RPMs and boost pressure I still can't see where that would create an EGTs problem any worse than with a turbo. As I understand it high EGTs aren't caused by excessive air flow into the motor - that actually results in cooler exhaust gas. Where EGTs get out of control is when you have high boost numbers and lots of fuel to burn with all that air. So if you properly control the flow of fuel through the pump & injectors, high EGTs shouldn't be a problem. Where the trick and the balancing act comes in is getting the right combination of boost and fuel delivery to maximize power without exceeding the safe EGT limit.
With a Paxton type supercharger the parasitic losses aren't nearly as high as with a roots type. The rotating mass in a Paxton is comparable to what is in a turbo - a compressor wheel. With the roots you have a pair of big honkin' meshed rotors with about a thousand times more rotating mass than the compressor wheel of a Paxton AND a set of gears to sync them - and that is where your most of your big parasitic losses come from - spinning those big hunks of aluminum. The other cause of parasitic loss with the roots is that it is a positive displacement air pump - meaning it is going to continue to push the same amount of air every revolution regardless of backpressure, and trying to overcome that backpressure is another form of parasitic loss. The Paxton, like a turbo, is a centifugal type of air pump and when there is backpressure the compressor wheel essentially "freewheels" - meaning almost no parasitic loss due to backpressure.
The only advantage of the turbo over the Paxton is that the turbo gets some of the energy to spin the compressor from exhaust heat and the resulting expansion of the exhaust gasses. But it still isn't completely free, because even the best turbo creates some exhaust backpressure, and we all know that diesels run their best and most efficiently with no exhaust backpressure at all. So there is still some parasitic loss pushing the exhaust gas out past the restriction of the exhaust turbine. The Paxton gets all of the energy to spin it's compressor directly from engine rotation, but it creates no exhaust backpressure at all. So it has a little more parasitic loss than a turbo - but not a huge amount.
The G-lader supercharger that VW put on their G60 Corrados, and Passats in the early 90s is about the best modern example of a factory installed belt-driven non-positive displacement supercharger I can think of. In terms of design the G-lader isn't exactly like a Paxton, but functionally it performs pretty much the same. In a Corrado with a 1.8 liter 8 valve engine a G-lader will boost the HP nearly 50% (from around 105 to about 150), and the energy "cost" is just a couple of mpg compared to the same engine without the supercharger. In fact, the 8-valve 1.8 liter Corrado G60 gets about the same gas mileage as the 16-valve 2.0 liter in my Jetta GLI, but it makes 10-15 more ponies - even though it has 10% less displacement and half as many intake/exhaust valves. And the two cars are built on the same chassis and very comparable in terms of curb weight and aerodynamics (the Corrado is a little more aerodynamic, but not much).
Although the compressor RPMs relative to the engine RPMs will need to be much higher with the diesel than with a gasser, that shouldn't be a problem because the compressor would still reach those same high RPMs on a gasser - the engine would just be turning at higher RPMS too. Getting it set up right on the diesel would require calculating the RPMs and boost (PSI) of the compressor and the air flow (CFM) at the HP peak RPMs of the smaller displacement gasser, and then setting it up the pulley ratios to spin the compressor at a rate that will provide the same boost pressure with the larger CFM displacement of the diesel at the diesel's HP peak RPMs. Not an easy set of calculations to do with any real precision, but not too tough to calculate a reasonably close estimate. It doesn't have to be exactly perfect to work - just reasonably close.
I think a Paxton would be a very worthwhile experiment on one of these IDIs, and a pretty easy install too. Getting the pulley ratios just right might involve adding a second pulley to either the vacuum pump or alternator to use two belts and step it up in two stages, but that shouldn't be an insurmountable challenge. The air conditioner compressor mount would be the logical place to put it with a minimum amount of fab work, but if you HAVE to keep your precious AC, then you could still probably fab the brackets to mount it below the AC compressor or vacuum pump.