Psiron,

Looked at the video and a couple of things stand out.

He only shows the machine doing rapids and he has set very low accellerations on the two main axes. This sort of negates his claim of 1000 IPM, it will average considerably less. As he states himself it only gets to that speed just before it hits the limits. Don't be lured into the rapids trap. Most of the cutting you do will be at or near feedrate speeds and I can't find a video of it cutting the pieces in the gallery. I still think that gantry will twist under cutting forces. Of course you can still get the sort of finish he has got in the gallery from a bendy machine, you just need to slow down the feedrate and reduce the cut/stepover/stepdown etc. It's a PITA though and much better to do it right first time.
BTW, speaking of limits, you really shouldn't use a limit switch as a physical stop as he seems to have done. On my machine the limit switch is set to trigger at least 25mm before the gantry hits a stop, the limit is a roller type switch and the trigger is a wedge of plastic. My gantry will actually roll off the end of the rack before it hits a stop so no mechanical damage can occur.

Someone else (Irving?) might be better able to comment on the long term viability of using belt drives over such large distances. 16 feet of belting per side is surely going to have some sort of backlash in it. I know small belts are regarded as "zero backlash" but at these lengths, ambient heat will surely have some effect plus the belt must sag as it expands with working. It might be no more than microns but it might be much more.I'm stabbing in the dark here but these seem like common sense questions to ask. I do like that he is driving both sides of the long axis from a single motor, you are not going to get racking that way but what if one belt expands a little more than the other?

It look a good sensible design overall and one that can be easily modified if things don't work out as you wanted.

Jeff.