On my ppc64le machine with 32 hardware threads, this speeds up OpenArena (1920x1200) from 7.2 fps to 8.1 fps.
As best I can tell, there is no documented reason for the current max being only 16, other than the commit message for 38a751cb saying that it was bumped from 8 in 2013 because someone reported a resulting speedup on a mailing list.