As a start the first thing you could do is to physically measure the difference in height between the center of the diaphragms although this is only a rough guide because the acoustic center may not exactly coincide with the actual tip of the dust cap. The speaker driver that is the most recessed in is the one you would use as a zero delay reference (usually the woofer). The other drivers that protrude above the reference needs to be delayed based on how much it protrudes. Lets say it protrudes by DX meters with respect to the woofer, then the time delay DT needed to be applied can be calculated by:-

DT = DX / c where DX is the displacement offset between drivers in meters and c is the speed of sound which 345 m/s at T=20 degrees C.

Once you calculate DT then you can calculate the delay in steps of sample rate interval needed by Audioweaver which in the case of the UP is 1/fsr or 1/192000 (roughly 5uS steps)

Therefore the number of sample delay steps you program into audioweaver is:-

sample delay = DT * fsr = DT *192000

For example say the delay offset between the woofer and midrange is 5 cm ie DT= 0.05 / 345 = 145 uS

sample delay = DT * 192000 = 27.82 rounded to 28 !!

Now, once you have come up with a first approximation of the delays, then to come up with a more exact figure will be tricky and it will be more relevant for the higher frequency drivers where the wavelength compares to the physical offset. So it is much more critical with the differences in offset between midranges and tweeters than it is for the woofers.