Linkage Disequilibrium (LD) is a well known tool in genetics.
Laurent Excoffier developed a test for the linkage disequilibrium between pairs of loci in haplotypic data such as STR haplotypes. It is part of his software package Arlequin.
I looked into this approach in regard to three sets of haplotypes within a 3 sub-clade: DF41*, Z305*, Z220*. The estimated age of these sub-clades is between 2.7K and 3.4K
The results can be summarized as follows:
Z220* haplotypes had 4 loci that were linked for circa 30 % with other loci;
In DF41* 6 loci could be linked to 30-50 % of the other loci;
in Z305* however every locus was linked to 40-100 % of the other loci.
Simple said: the Z305* loci changed very much in unison between the haplotypes of the set;
the DF41* set showed some harmony in their development;
Z220* is clearly a discord concert were everybody plays its own tune.
Sets with similar GD's but different LD's are clearly different kind of beasts.
LD can help us to characterize in qualitative way haplotypes within a sub-clade were GD, estimated age, variance at best give a quantitative indication of the subclades.