"His Monte Carlo analysis shows sample RSD to be most efficient ..."
I think there should be a caveat here that "most efficient" was measured assuming circular groups and that there were no flyers. Mean radius would be less sensitive to a flyer than measure based on a standard deviation.
Corrections To Paper
Ref 3 should it be Stephen not Steven (?), and Mosin-Nagant not Moisin-Nagant(?)
Haven't seen article so don't know for sure about Moisin - that could have been the way it was spelled.