Tcksample on dispersed streamlines

Hi all,

I wonder how stable is to sample metrics like FA via tcksample along streamlines in tracts with high dispersion or curvature, e.g. the uncinate fasciculus?

Thank you for your insights,
Dominika