optimal diatonic tuning again

It turns out (according to Mandelbaum) that Wesley Woolhouse used the
criterion, RMS (major third, minor third, perfect fifth), in his 1835 study.
He finds that 50-tone equal temperament is nearly optimal, but advocates
19-tone instead for its simplicity and nearness to the optimal tuning. As I
recently proved, the actual optimal tuning according to this criterion is
7/26-comma meantone. 31-tone is closer than 19-tone to this value, and
despite the simplicity of 19-tone, I would definitely choose 31-tone if I
had to choose one, since the possibilities of expanding beyond diatonicity
in a musically meaningful way (including 7-limit and 11-limit harmony and
Arabic scales) seem far more promising in 31-tone than in 19-tone.