Having a character encoding for shorthand kind of defeats the purpose of performing paleographic analysis on handwritten shorthand, because in order to encode, say, a Pitman or Gregg "d"' you must have already identified it as a "d". An encoding that described shorthand strokes in terms of length and direction (sort of like dance notation) might be more useful for this.
Do you know how the French stenographic system works? It's not encoding letters and not even phonemes. It's really more like an abjad (like Arabic), with some characteristics taken from Brahmic scripts too (the way some phonetic consonnants like L and R are handled, as they are often used in association with other consonnants), and with ambiguous consonnantal phoneme or groups of consonnants, with optional diacritics for vowel sounds, and special symbols for common terms or prefixes/suffixes. There's no such symbol that encodes a "d" letter.
There's no exact one-to-one correlation between Latin and this stenographic system (here I speak about the Foucher's stenographic system, that was officially tought for decennials with one of two methods: a complete one, the most concise with more complex joining rules, and a simplified one, with more regular symbols, and much less exceptions).
There's abolutely no way to make a coherent and correct representation of the script using a transliteration to Latin, or even to IPA.
I note that this is not an orthographic system, but a specialized notation mixing phonemic/phonetic (like IPA) and semantic items.
It's true that this script requires precise relative length and directions to be readable, and that the script supports very little glyphic variation as permitted in Latin-like scripts, but other encoded scripts in Unicode also have this requirement.
I don't have the story and theory of the script, but I think this could be in some research papers published by Foucher on the phonetic and semantic structure of the French language...
The rare documents that my mother has left about it are undecipherable by me, and some book fragments about it are not enough complete to make any proposal or theory about how it works, and how it could be learned easily (in a course less than 1 year) by French secretaries.
If only I had the training course materials that was used in the 60's...
