Some improved poptimal generators

I wrote code for least-cubes optimization for linear temperaments, and used it to improve a few of the values. I was surprised by how often the least cubes value allowed me to slightly enlarge the interval of known poptimal generators, but for the most part this made no difference. It did, however, in three instances; and I was especially happy to find that the 94-et is poptimal for schismic.

Quartaminorthirds [9, 5, -3, -21, 30, -13]

[1, 16/247], ==> [1, 9/139]

Supermajor seconds [3, 12, -1, -36, 10, 12]

[1, 59/305], ==> [1, 53/274]

Schismic [1, -8, -14, -10, 25, -15]

[1, 134/229], ==> [1, 55/94]