NavList:

Robin Stuart you wrote: I wonder whether the association of the UPS with its "Mercator" misnomer might have arisen because parallels and meridians form a rectangular grid in both.

I think there is a lot in that.  If you know exactly where you are and just want to test the accuracy of your celestial, or if your using a computer programme with your DR position as Z intending to transfer the fix to a larger chart, then a line up and a line across a piece of plain paper will do, although using graph paper makes a tidier job of it.  Your known or DR position is where your lines cross.  Azimuth is plotted from here, and mile is the same in all directions. 

Once you start to bring lat and long into it, either because you want to reduce LHA to the nearest degree so you can use tables; you want to include early or late shots; or you want a lat and long straight off the chartlet then, unless you live near the equator, you need a bit more thought, because a minute of longitude is shorter than a minute of latitude.  Forgetting Polar Regions for the moment, life is going to be easiest if the meridians go straight up and down the page and the parallels go across the page. 

If using graph paper, the temptation is to keep the parallels a regular number of graph paper divisions apart, but then you need to shrink the distance between meridians so that a mile is still the same in all directions. This makes plotting an assumed position so you can use tables more complicated, because it won’t be a whole number of graph paper divisions.  Better to keep the meridians 60 graph paper divisions apart and stretch out the minutes of latitude, which you can do by multiplying a graph paper division by the secant of the latitude (check at 0N x by1, at 60N x by 2) or do the clever stuff with your compasses.  This is why your chartlet start to look like a typical Normal Mercator chart at a similat latitude.  Purists please note I only said “look like”.  DaveP