A couple of observations:
Maybe worth mentioning that many monuments that are "clearly" related (at least in terms of date/type/proximity) aren't necessarily all visible from each other, though some might be visible from some of the others. Some of the South Wales monuments may be not visible from each other due to a rise in the intervening ground. Postie and I saw good examples of this at Y Pigwyn, a few miles NW of the Maen Llia monuments.
Conversely those same monuments might all be visible from a different (often higher) vantage point. The summit ridge or western slopes of Fan Llia may be the "best" place to see the Maen Llia monuments from collectively, for example.
I would also chuck in that contemporaneity isn't always essential for monuments to be "related" either. Plenty of long barrows have round barrows built close by that evidently have a relationship with the much earlier monument (even if we don't know what that relationship is, exactly).
Even with good 5m contours, a 1/25000 map doesn't show you the folds in the land all that accurately. Here's a site that looks straight from above/on the map, but appears to curve on site:
http://www.themodernantiquarian.com/post/89397/down_tor.html
In any case, a surveyor won't lay a tape between two points on a slope and then use that measurement alone as the basis of the length of line he draws on his 2D map. A line measured from his 2D map will be shorter than the length measured on the ground, as he would need to correct for the slope. http://en.wikipedia.org/wiki/Tape_correction_(surveying)#Correction_due_to_slope
Can't say I've considered equidistance at sites before, personally. Something new to think about!
I would tend to agree (based on nothing more scientific than visiting sites without measuring anything at all) that topography plays a bigger part than equidistance in placement of monuments in proximity.