[css-speech-1] CSS speech, a first step toward standardized TTS APIs?

[LaurenceZaysser]

PROBLEM
So far, every TTS engine has its own proprietary API to interface with screen readers, and eventually with web browsers, eBook reading softwares (to offer a dynamic audiobook service like in Thorium, Librera, etc.), and other applications.
So far, free TTS engines are provided on various "reading machines" (smart phones, tablets, personal computers, digital book reader). But for a better user experience, blind people prefer paid TTS (if they can afford them) to read text aloud, instead of the free ones (of lower quality).

So far, screen readers interface with TTS from the OS side and connect to the user's preferred TTS. No web browser, nor eBook reading system is able to do this. Why ? Because no standardized TTS API does exist. Though, the need for such an API is getting more and more important, in order to directly interface various applications (instead of OS) with various TTS engines.

SOLUTION
CSS Speech could be a very good base for defining a standardized TTS API. Please keep this in mind when you define CSS Speech.
Furthermore, in this perspective, what should it be? The smallest or the greatest common denominator between TTS engines?
Your answer will be ours... ;-)