How hard would it be to... (full build imaginings)

[Nexus]

Borrowed from cosworth MX-5 good for 175Kw atw before you apply turbo :
mx5-nc-supercharger-rweb1_668_general.jpg

Also simpler install than suggested external recirc. not wure you could do that with this. blower and intercooler in-manifold.
Presuming it fits our MZR variants?

The other thing that occurs to me.....brand new MX-5 cosworth comes out about similar coin to a 2007 BMW Z4M which is pretty high on my achievable-fun-performance wish list.

but back to the MPS6 avant...should this be installable, turbo is applied as per normal, turbo TC bypass before sc+manifold. easy. Install + tune.

[Nexus]

Reseached a bit more : This is definately doable. The cosworth kit fits to a different head with different ports, so there's no bolting to an mps, but the belt configuration is shown in image :
Cosworth_MX5_SCKIT_03.jpg
Is valid for both the MX-5 and MPS MZR engine variants. There's your belt layout.

But the eaton assembly is a non-starter, so it comes to careful placement of a sprintex S150 @ 2.5-->2.6:1 drive ratio, and a custom intake manifold to accomodate the supercharger. ditch the variable intake runner assembly for a simpler SC force fed manifold, and it's a goer. It's a tight squeeze but it's achievable.

Three configs possible relating to throttle plates and recirculation:

(1) Place SC after TP - No external recirculation.
(2) Place SC before TP with secondary TP and (high boost) sealing valve controlling external recirculation (via turbo).
(3) Place SC before TP with secondary TP before SC plus tertiary TP and sealing valve controlling external recorculation and providing fine control over SC boosted intake zone pressures, and the ability to operate the SC below atmospheric pressure, reducing pumping losses when desirable for the circumstances.

Option 1 minimizes pumping losses operating the SC ina vacuum behind the TP when off throttle.
Option 2 has highest pumping losses operating at a minimum of atmospheric pressure, but recovers some, all, or more *unlikely* losses through exhaust scavenging.
Option 3 allows the SC to operate in both modes and anywhere in between, both minimizing pumping losses and maximizing recovery of pumping losses via exhaust scavenging by varying the airflow through the SC to suit more varied conditions.

Option 2 good for track. Option 3 is what a manufacturer would want in terms of flexibility and energy efficiency.

To implement, an inteceptor or standback would be required to operate the secondary (/tertiary) throttle plates as a form of boost control. Might as well have it handle the turbo too, but the tuning would be as per normal with the exception that we'd be paying especially close attention to the bottom end. Option one could be implemented without an inteceptor/xede/standback however TP would close further on idle etc...creating a high-vacuum zone that might be desirable in track environ for brake booster - I don't image an extra -3psi hurting your brake booster, but I do imagine it improving the booster's effectiveness. Someone correct me if that's a bad idea.