NavList:

Ed wrote: "Centering non-full moons is challenging"

The non-full Moon is challenging for all bubble sextants, I believe. (With adjustable bubble size it might be somewhat simpler.)

I developed a process for the my A-12, which is suprisingly accurate with crescent Moon as well. (Most probably I reivented the boiling water, but I could not read about it before.)

I measure the selected (upper or lower) limb of the Moon against the top AND the bottom curve of the bubble. I use the indirect view, so the Moon looks upside down. I match the selected limb to the top curve of the bubble, make a mark on the disc, than match the same limb to the bottom curve of the bubble, and mark it too. The bubble must stay centered. In a two minutes interval I make roughly twenty alternating measurements (ten at the top and ten at the bottom curve of the bubble). The alternating pattern is important, because the Moon moves. Reading the averages of the two series I got two altitudes, one at the top and one at the bottom of the bubble. (the difference is actually the size of the bubble: 60-90 moa). Taking the mean of the two altitudes I got the Hs for the selected limb of the Moon. Correcting for index error and ignoring the dip, I use the Moon corrections (refraction, parallax, augmentation) as per Nautical Almanac.

It is more accurate than centering the Moon, because matching the limb with a curve could be more precise. In addition this is basically two measurements averaged, adding some extra accuracy. 1-3 nm error in LOP for a crescent Moon is achievable with a handheld A-12 on ground.