I've been looking for some data on estimated drift from this, but no specifics. That said, since it's looking at wave properties of cooled atoms, I'm assuming

I've been looking for some data on estimated drift from this, but no specifics. That said, since it's looking at wave properties of cooled atoms, I'm assuming we're talking "wavelength" accuracy as opposed to more conventional inertial nav systems. My point perhaps is that though this is "nothing new", it's (probably) way more accurate than anything before it. So the "pain points" of recalibration and drift are extremely minimized.