top complexity

orthogonalization.

since it doesn't matter which comma basis you choose, you can always
choose a basis where the commas each involve n-1 primes (this is
probably one of the matrix reduction or decomposition methods matlab
is happy to do). then it's trivial, up to torsion, to express the
complexity in terms of the complexities of the commas, since the
relevant {length, area, volume . . .} measure will just be that of a
rectangular solid . . . or do you have to iteratively cascade down
the dimensions (brain foggy . . .)?

The orthogonalized basis for pajara would be {64:63, 50:49}. The
complexity should be the product of the complexities of these 3,
with the suitable "dimension boost" factors = log of the remaining
prime. yup?