For me, I'm curious what the main focus of the camera (assuming it materializes) will be. Would it primarily be a studio camera? If so, it is not of much interest to me. But if they make a small and light (relatively) rangefinder that handles well in the field, that really gets my interest. The X-T1 is about 40% lighter than a D7200. A Fuji Medium Format Rangefinder could end up similar to a Nikon 810.
This is the big sticking point. The advantage of the Pentax 645Z is that you can work quite fast with it considering it is a MFD camera. It still forces you to slow down compared to a DSLR though.
The only reason the Pentax 645Z is quite fast for an MFD camera is because it is an SLR design. A mirrorless design is inherently slower than an SLR design, so that if Fuji had made a MFD version of the X-Pro1 with the Sony 50MP cropped MFD sensor, it would have been intolerably slow.
The question for Fuji is whether mirrorless technology has advanced enough that a mirrorless MFD camera can be made to shoot fast enough to keep up with the Pentax 645Z. It's all very well if the GX-Pro1 is more compact and portable in the field than the 645Z, but if it is so slow it might as well be a studio camera for shooting with strobes, its inherent slowness would undermine any size advantage. Keep in mind that the 16MP X-Pro1 already forces you to slow down, so imagine how painfully slow a 50MP GX-Pro1 might have been if released 1-2 years ago.
The other thing is that the competitive price of the Pentax 645Z is said to be due to the fact that it shares lots of parts with their DSLR bodies. Fuji too need to have a GX-Pro1 share lots of component parts with the X-Pro2/3 body. So in many ways, the development of their MFD system is dependent on that of their X-system. If the X-system was successful then that would fund the R&D costs of their MFD system, just as the Instax system funds the cost of the X-system.
Is the mirrorless technology in the X-Pro2 mature enough now that it can be upscaled to MFD proportions? Or should they wait to upscale the X-Pro3 to MFD proportions, and deliberately design the X-Pro3 so it can share manufacturing parts in common with the X-Pro1 to reduce costs?