equal steps for melody

The 5-limit diatonic scale has two step sizes -- or is it three? In just
intonation, it's three:

9/8 10/9 16/15 9/8 10/9 9/8 16/15

and Bill Alves might argue that this is just how the diatonic scale
should be, melodically or otherwise. However, compare the 5-limit
diatonic scale in 12-equal:

2 2 1 2 2 2 1
(2 step sizes)

with that in 22-equal

4 3 2 4 3 4 2
(3 step sizes).

The 12-equal version sounds better melodically, even though 22-equal has
consonant intervals closer to those of JI. I think this is because
having three different step sizes is awkward melodically (in this case),
and in JI the 9/8 and 10/9 are, as Ben Johnston would say, close enough
to sound like melodically equal steps.

Note that while the 5-limit led Ben Johnston to a 53-near-equal JI
scale, the 7-limit led him to a 22-near-equal JI scale (which I read
about in Perspectives of New Music). Anyone familiar with other
near-equal scales Johnston used?

Bill Alves' was my favorite piece on the tape swap, and I think its
strengths are rhythm and harmony. The harmonic basis for the melody was
strong enough that even the few awkward melodic passages can sound
somehow "right," though in a very exotic sort of way.