I am. Does your original TOD map contain NaNs? Feel free to send me the files
mnireg/t_t1_2_mni152_mr.txt mni152_down.mif todmap_norm.mif maps/todmap.mif and the original TOD map (I think one of `maps/todmap.mif, rois_and_tracks/todmap.mif). The interpolation gif shows that something odd is going on.
In any case, I’d second what Thijs suggested, warp the .tck file (using the inverse transformation, see here) to template space and generate the images there should, from what you describe, solve the normalisation issue. This doesn’t solve the open questions, so if you can share the data, I’d be happy to investigate.