Hi,
I am just studying the FreeSurfer open source code (obtained from the read-only CVS, “dev” branch checkout instructions). It appears there is a slight mistake in mrisComputeOriginalVertexDistances. The following code computes the difference between v->y and vn->origy, where it probably should use v->origy as well.
mrisurf.c:4093: xd = v->origx - vn->origx ; yd = v->y - vn->origy ; zd = v->origz - vn->origz ;
Andreas