Fuji has a small site where they are writing up tech info on the GFX. So far they have covered the sensor design and G mount. Here is a link to the Fuji write up on their version of the Sony 50MP. Appears that Fuji may have made some changes but without a camera to test, hard to really say.
Thanks, Paul. Unfortunately, this is written in such a way that I can gain almost nothing useful out of it.
What I think they are saying:
The FWC has been increased by a factor of 2^(1/3) = 1,26.
The effective fill factor has been reduced from nearly 100% to something less. This will increase MTF50 and also increase aliasing.
They have increased the sensor scan speed during AF acquisition from 130 frames per second to 200 fps. If they refresh the EVF at this rate, that should also reduce EVF latency, but they don't say whether they do that or not.
There have been some packaging changes that I can't figure out: "Of course, the sensor itself has the surface size of 43.8x32.9mm, but the shape of the sensor unit is different for the GFX 50S. Typically the sensor is mounted to the silicon circuit with terminals placed on both sides of the sensor. But in the case of GFX, the sensor is mounted to the silicon circuit with the terminals placed on the back side of the sensor. "
My problem with this statement is that the sensor is itself a silicon circuit. Are they saying that they are using a stacked sensor? That would be news to me. Or is it as simple as saying that the package terminals are located behind the sensor? That wouldn't be rocket science. But for that to be true, the package would have to be the thing that they're calling the "silicon circuit", and that doesn't make any sense: there's no silicon in the package, at least to speak of. If it's just about the package configuration, there's a downside to their approach: it will not be easy to attach a heat sink right behind the sensor.
There is no mention of different CFA dyes, which I would expect would be the first thing they'd change, since Fuji has a take on color.
They have come up with a term that is new to me: photic saturation. I'm guessing that it is FWC divided by QE, but I don't really know.
At least I'm confused at a higher level.