The VMS as a possible chain encryption ( Mod 23 ).

[RadioFM]

Although handy, I think we should really make an effort to not hold character entropy as the sole, go-to metric for the VMS. With verbose encoding, conditional character entropy can be driven (almost arbitrarily?) down to VMS levels and n-gram frequencies plots skewed. And that's not even accounting for null characters that can skew and mask stats even more.

Here are some rough stats with bim3w's ciphertext and a verbose (reversible) remapping of characters to keyed numerals, like "A" → "V1", "B" → "V2", etc. I'm disregarding all word and linebreaks and remapping n to i for demonstration purposes.