I would think you could find equivalent star patterns for most formations, even ones with 5 or 6 points of reference, if you look at the stars as a random scatter, free of the constraints of the traditional constellation patterns. That being so, it's always going to be arguable whether it's chance or not, although it would make the basis of a great book. I guess you could do the same with Rock Art (or crop circles!!!) and who is to say whether it's valid or not.
The only thing that would tend to be convincing is if a particular star pattern appears in quite a few monument formations.