I'm aware of some studies that modify Milroy's index, but all of these indices are rather arbitrary. I'm not actually aware of any variationist studies that make use of things like PageRank, degree centrality, closeness centrality, etc., i.e. centrality measures that have actually been thoroughly developed in social network analysis.
I'm using such measures in my thesis, but I don't want to assume that they're simply never used in sociolinguistics, and if they are used, I want to be able to see how they've been used elsewhere and compare results. If anyone can point me to some relevant studies, it would be much appreciated.