I need to step away from this for the evening. I have a history as a programmer and I still cannot comprehend in which orders these levels ensure that something is said versus when it isn't.
Based on Tony's description these seem to be "skip speaking the item levels," with none meaning always spoken, while what I'm getting from your description is the opposite.
It would be so much easier for me if these were numbers rather than English descriptions. At least then the hierarchy order would be clear.
In any case, my main concern at the moment is if one has a series of symbols that compose a word, but those symbols are not letters, and you want to be certain that these symbols are always spoken each and every time they're encountered, what does one set the level to in order to ensure that. They'd never be passed to the synth since it's the symbol name (replacement text) that will need to be spoken.
Brian - Windows 10 Home, 64-Bit, Version 1809, Build 17763
A great deal of intelligence can be invested in ignorance when the need for illusion is deep.
~ Saul Bellow, To Jerusalem and Back