Glad it worked. What was it that made it work eventually, so I can add it to the github page? Since the system by default blocks file writing for Personal Voices, when a personal voice is used, we cannot write the buffer directly to file, so the library lets the original say command to play the output of the audio in order to record it and write it to file instead. So the playback delay is required...I can only mute the audio output, but this wouldn't change anything in terms of delay.

1. From Terminal, run: sudo nvram boot-args="-arm64e_preview_abi amfi_get_out_of_my_way=0x1"

At this point I usually find that Live Speech via say -v MyVoice ... is producing nothing but silence. I pull down the Live Speech menu, enter a word, and it produces silence. Enter a word again, it says it in the default system voice. Enter it a third time and it says it in my voice, and now the say -v command works correctly.

(Steps 1-3 are also needed for authorize_terminal to run)

Thanks. Yes, for the last part, we can modify the code to store files in different formats as well, that was mostly a proof of concept.