Starting with number 7), tell me where I go wrong.

The moon is 0.5 of a degree wide (give or take) when seen from earth.  This means that if you drew a moon next to a moon next to a moon all the way around its orbital path you could draw 720 moons.

You say that the navigator necessarily has an error when finding the centre of the moon.  Could you expand on that?  If I had, say, a circle cut in a piece of metal that (give or take) represented the moon, then by positioning the moon within that circle, the plumb line would cut straight through the centre and down...so one could see the stars in another cut out, maybe one with degree markings--even a grid of some kind (fine hairs stretched tight?)

So you are saying that the human eye-moon relationship has a necessary innacuracy--error-of 1/27.4th of the width of the moon?

8. and 10) I don't understand "Parallax error".  Would it help if the almanac ran through the entire sol-lunar cycle, so accuracy would remain constant as the moon made its various deviations from the hypothetical perfect globe moving in a perfect circle?  If one knows that on day X there will be a human-viewing error of Y%, then one could use that known error in the calculation (like sighting an inch above the target because you know that given the relationship of sight to barrel to bullet, you'll hit the target?  Am I way off?)

9. had me thinking of ice ages again!  Before the last great melt, the ice came down to 45 degrees north, so we can imagine people working in a 45 degree area north (and south?)  Is that enough for serious problems to arise?