I think that in arguing telecentricity (and please lay off the real or simulated obscenities), you're assuming that a simple sensor with telecentric lenses is the only way to go. Maybe it is if corner performance is all you want (I'm not sure that any 90-degree image has risen or fallen based on the last 5 mm of the diagonal - but let's assume it's a drop-dead issue). To get that corner performance, you are working with a 16Mp limit (thus far), a big size, and the quirks of Fuji prime design (such as super-compressed DOF scales that almost make you question why they're there). One consideration peculiar to Fuji is that as a proprietary system, there is zero choice when it comes to optical or handling characteristics of lenses. Or bodies - and to bring the discussion full circle, a closed ecosystem means that no matter how good the optic is, you have to consider the baggage.
And on IQ, have you actually run the comparisons? I have tested the 21/4.5 Zeiss ZM (which is reportedly the worst lens for a digital camera) against my 14/2.8 Fujinon (which is Fuji's widest prime lens, and probably the best 90 degree lens made for digital). There is not much in the way of noticeable corner degradation with the ZM on an M typ 240, and by the time you downsample to match the X-Trans sensor's resolution, it's gone. Color shifts are sometimes an issue with wides, sometimes not, but the software for correcting this (even on the ZM) is faster than Adobe's straight up conversion of X-Trans files (I have a 4Ghz Retina 5K with 32Gb of RAM, and it still takes up to 5 seconds to render X-Trans files). What you can't correct, even given unlimited time, is the fact that a X camera with the 14 and a body won't fit in a coat pocket.
On the other hand, on a Leica, I mostly use a 21-35mm M-Hexanon Dual. Though a big lens for a Leica (it's actually the size of a 50mm f/2 except for the filter mount), I'm pretty sure is still smaller than my XF 14 (I'll measure it tonight), covers 24x36, acquits itself well at 24Mp, has two focal lengths, no corner or color shift issues, and actually has a usable DOF scale (actually, two of them).
A lot of this will boil down to individual preferences, but if Matt is right and this is really a DSLR-competing system (and having looked at Fuji's X-Pro press releases, they only say "comparable to a DSLR" in image quality), then it is not surprising that the X system will approach DSLR lens sizes. Maybe we all toiled under the misapprehension that because it looks like an RF, the X-Pro1 should be compared to one.
Dante