MediaStream, ArrayBuffer, Blob audio result from speak() for recording?

Hello,

What would be necessary to implement an optional parameter to `
window.speechSynthesis.speak()` which when set would return an
`ArrayBuffer`, `Blob` or `MediaStream` of the rendered audio?

Located this Answer https://stackoverflow.com/a/18903582 which discussed a
similar feature. Searched the mailing list though could not locate the
discussion. Am not familiar with c++, though also browsed the chromium
source for chromium implementation.

Can you linked to the mentioned precious discussion? Are there any current
discussions as to implementing a method to get the raw audio output to
JavaScript?

Have been able to record the output using `MediaRecorder` and
`navigator.getUserMedia()`, though the recording contains sound artifacts.

`meSpeak` http://www.masswerk.at/mespeak/ implements a `rawdata` option.
How can this feature be implemented in chromium in general? Or a custom
chromium build?

Regards,
guest271314