One comment I have on this is that doing so does not automatically result in a circle or curve. You could draw a straight line on the ground and walk along that marking the spot where the alignment is apparent. This could actually explain some stone rows that don't point at anything, but are in places where obvious 'sighting' points lie in obvious directions.

I don't know of anyone else who's considered rows and alignments with this in mind. I've started to study a few stone rows in this manner and have some ideas about a couple of them, but more time is needed.

To make a curve or circle you would have to tie another pole to the positioned one and then move with the string taut to get the spot for your alignment. If you tried to do the curve thing by eye you would almost certainly end up with an elipsoid thing. But many stone circles have equal spacing and the technique above wouldn't create that. Also, it can't be applied to the vast majority of Irish stone circles as they conform to patterns - 5, 7, 9, 13, 17 stones etc. Maybe some of the Scottish or Cumbrian ones fit the format?

Haven't read the link in detail, Nigel, but will later.

I think that, for what you are suggesting to work, the setting positions of the moon as seen by the observer would have to move from day to day in a way which corresponds with what the moon is actually doing. But the observer doesn't necessarily see what the moon is really doing because of the effect of refraction.

Here are some quotes from Douglas Heggie's "Megalithic Science".

"[Refraction] makes the apparent altitude of a rising or setting body higher than it would be in the absence of the atmosphere by an amount which can exceed the apparent diameter of the full moon. ....

Atmospheric variations vary from night to night, and the resulting variations in refraction would have limited the usefulness of very accurate orientations even in megalithic times. In other words, unless conditions then were much more stable than they are now, these variations impose a limitation on the accuracy with which a fixed megalithic orientation could have been used to record a fixed astronomical position. ...

Refraction is particularly subject to variation just above the level of the ground ... [distant foresights] tends to make refraction stronger than it would be otherwise. Perhaps it also implies that the declinations defined by these lines can now be determined only with added uncertainty, and that their possible use in megalithic times was rendered particularly troublesome."

As an example, he adds
"... refraction changes would have prevented the use of the accurate natural foresight at Ballochroy, Argyll from determining the day of the solstice since they much exceed the daily variation of the declination of the sun at the solstice."

In other words, what I'm saying is that refraction is not simply a problem to modern day surveyors trying to measure and replicate what was seen in the past. It was also a problem to those trying to observe and record movements in the past if they were trying to achieve highly accurate recorded positions which they could return to a year later or whenever and expect to see the moon or sun in exactly the same place as it was when the recorded position was set up.

I'm not anti-alignment; just unconvinced by many of the claims for very high accuracy.