If you have the drafting and flow dynamic modeling software, a CNC machine and a flowbench, all you need are the dimensions of the OEM manifold as a starting point and you can try several iterations until you see good improvements.
I think the real bottlenecks will be the exhaust cam -- which is not variable -- and the exhaust manifold. You might be able to work around the exhaust cam, but I suspect the IM will only be really worthwhile for folks with headers.
Assuming you can work around those issues, the last major hurdles will be space and cost. Fitting different runners under the hood without other modifications ( e.g., different strut brace) will be tricky.
Bigger diameter runners should give more overall torque, You might play with stepped diameters too. I believe shorter runners will net a peak torque at a lower RPM, while longer runners will make peak torque higher up.
Last major obstacle is cost -- it would be nice if these could be mass produced for not much more than a grand, and less would be ideal. so sacrificing some shiny/pretty for power seems like a worthwhile trade.
I think the fundamental problem with building an IM for the Z has been the approach; I think all other IM attempts have been based on adapting an existing IM from other applications rather than experiment with developing a totally new unit using a design where the OEM 370Z IM is the starting point so you have a clear benchmark to work with.
If you have the materials, the tools, and the knowledge to develop something like this, go for it
