For context, as a software developer and Mac OS user who also happens to daily drive a screen reader, I seriously doubt whether you could implement a third-party SR on that platform.
It seems that third-party software, even software with accessibility permissions, doesn't work on password screens (and probably in a few other similarly-secure places), and you need those to be accessible. Not to mention weird places like system recovery, which (for very obvious and understandable reasons) does not allow 3rd-party software at all.
I guess you could use a third-party SR for most of your system and then toggle VoiceOver on when accessing the secure parts, but that would get very annoying very quickly.
There's also no 3rd-party access to some speech-related features, like the higher-quality neural Siri voices. You'd also need APIs for things like automatically being informed of incoming system notifications to read them as they come in (which the first-party VoiceOver does), and those don't seem to be available at all.
It seems that third-party software, even software with accessibility permissions, doesn't work on password screens (and probably in a few other similarly-secure places), and you need those to be accessible. Not to mention weird places like system recovery, which (for very obvious and understandable reasons) does not allow 3rd-party software at all.
I guess you could use a third-party SR for most of your system and then toggle VoiceOver on when accessing the secure parts, but that would get very annoying very quickly.
There's also no 3rd-party access to some speech-related features, like the higher-quality neural Siri voices. You'd also need APIs for things like automatically being informed of incoming system notifications to read them as they come in (which the first-party VoiceOver does), and those don't seem to be available at all.