[GRASS-user] r.basins.fill usage

[M S]

Upon first glance, r.basins.fill seems like a great tool to generate
subbasins within a watershed.  Although I must not be understanding
something obvious and essential.

If one has to digitize in the internal ridges in a given watershed,
why run the ridge input through the module if one has essentially just
delineated the subbasins ridges?

Is there an advantage to using a smaller threshold on r.watershed to
get smaller basin delineations (ceasing when thin strip basins get
introduced)?

Thanks for any feedback,
Mark