The X100 VF has a larger FoV [almost a 26mm equivalent] than the taking lens, enough to provide margins. [This is also the case with some RF cameras, from Leica's to Bessa's...an equivalent situation when using a 50mm on an M6.]
The X100 taking lens frameline is a rectangular graphic projected onto the OVF...said to be 90% of its actual coverage. [A Leica M uses illuminated edge framlines, with only 85% coverage at best.]
The VF framlines shifts toward the taking lens in closer focusing...compensating for view parallax; so is in the X100 and most RF cameras. [Some RF cameras only provide shifted frame corner markings.]
[BTW, most SLR's, except Nikon F, F2..., do not provide 100% coverage either. What is not seen in the VF is not seen at all until after the film is processed. Manufacturers like to excuse themselves by saying its VF coverage matches a cardboard slide mount......]
I modified the lens mount in my CV 40/1.4 to bring up the 35mm frameline, which is now very accurate in my ZI...at infinity.
A simple and most reliable way to check all that is to place a piece of matte film or ground glass on the guide rails, set shutter at B, and compare what is imaged on the matte film and seen in VF. [This method can also be used to check lens hood cut off......]
Meanwhile, the X100 EVF sees exactly what the sensor sees, and 100% accurate. In fact, if you compare the OVF to EVF views in Fuji's web site...you know
that lone tree view...you will find the EVF view ~11% larger than the OVF in coverage and spilled over to the fixed-position graphic [EV compensation on the left, and distance scale in the bottom.]
http://www.finepix-x100.com/en/x100/hybrid-viewfinder
I am simply surprised that many never seem to have read and thought about what Fuji has to say...but opinions abound.