Appreciate the effort you must have put in, but until AI learns to speak other versions of English, it's a bit of a waste of time, imo.
Technological progress seems to think everybody are American. I once saw a clip from the news when Siri first came out. A reporter went around on the streets of Aberdeen in Scotland and asked people to use Siri as speech to text. It was way off and the only one it worked for was somebody, who moved there recently, so he didn't speak Scottish. Neuro-sama suffers from this too. She has American pronunciation, but the more her large language model learns from her British creator, the more distinctly British her choice of words becomes. For the few of you, who still doesn't know, Neuro-sama is the first (and only?) successful twitch AI streamer.
While I do get your point, "waste of time" is stretching it. While perfection is a worthwhile goal, the real question is if the change makes it better or worse, meaning is voiced better than unvoiced? Maybe the answer to that depends on who you ask.