Würschmidt's rational systems; diesis definition

The holy grail for "reasonable" Fokker periodicity blocks I had hoped to
find in Mandelbaum's dissertation turned out to be a bit of a
disappointment. This is the passage I remembered, pp. 39-40:

"W�rschmidt, in his theory of rational tone-systems, has made one of
the most penetrating analyses of the qualities and possibilities of the
various equal-tempered systems within the context of the web of fifths and
thirds. He is particularly concerned with two kinds of minute intervals
which he considers particularly important to tempered systems. The first he
calls defining (definierend) intervals. These are the very small intervals
in the web of 5ths and 3rds which are made to disappear (made equal to zero)
in the process of building a specific tempered system. The syntonic comma,
for example, is a defining interval in 12-tone temperament, as is the major
diesis. The second kind, W�rschmidt calls constructing (konstuierend)
intervals. These are the intervals each of which is equal to one unit of a
tempered system. The constructing intervals in 12-tone temperament are the
various kinds of semitones such as 16:15 and 25:24. It is W�rschmidt's basic
contention that a rational system is one in which the largest defining
interval is smaller than the smallest constructing interval. The rational
systems, according to W�rschmidt, are 5-, 7-, 10-, 12-, 19-, 22-, 31-, 34-,
53-, and 118-tone temperaments."

What I didn't remember was the footnote:

"W�rschmidt does pose a limit on the expanse of his system for, if he did
not, other small intervals far removed from musical practice would interfere
with his calculations. He chooses as limits 11 fifths and 27 thirds to
eliminate the very small discrepancies from the octave created by the 12th
fifth (the Pythagorean comma) and by the 28th third (which exceeds the ninth
octave by only a few cents)."

Mandelbaum and/or W�rschmidt call 128:125 the "major diesis," contradicting
Keenan's recent definitions. Is there an accepted convention on this?
Manuel?