Media audio to Bluetooth mono device (device normally used for phone calls)

I will use a car Bluetooth phone profile as an example, but the same applies to normal Bluetooth mono (SCO) headsets.

For now here are the choices:

1)  Constant connection to car audio (no car radio will play):  In SoundAbout, you set the "Media audio" setting to "Bluetooth (mono)".  This will send all phone audio, from the phone to the car radio by making the phone (and car) think there is a phone call.  The audio is only voice quality.   Notifications most likely won't be sent based on the phone's audio policy.

2)  Dynamic connection to car audio (car radio will play until phone audio sent):  In SoundAbout, you set the "Media audio" setting to "Bluetooth (mono)(dynamic)".   You can also set "Media audio" setting to "Let app decide", and then go into Bluetooth Behavior and check the "(Dynamic) Audio usage" setting.  This will also send all phone audio, from the phone to the car radio by making the phone (and car) think there is a phone call.  The difference is that this setting has the phone signal disconnected until it detects audio.  Once it detects audio, it connects to the car "like a phone call".   The audio is only voice quality.   Notifications most likely won't be sent based on the phone's audio policy.  If this setting is used for Navigation, you could lose a second of directions due to the time it takes to connect; which is why I am developing #3.

3) A new setting is called Text-To-Speech behavior with another setting called Audio output (like Media Audio):  You can map any TTS audio to various audio outputs, including Bluetooth (mono).  To do this, I have to create a TextToSpeech wrapper called SoundAbout TTS.  Within the setting of SoundAbout TTS, you can select any "real" TTS engine on the phone.  This will allow TextToSpeech requests to be capture by my app.  I can then switch the audio routing while holding the TTS.  Once the audio is re-routing,  I can let the TTS get translated into speech.  This way you don't miss anything.  I have this working, but Google Maps/Nav is sending me ALL the directions at once instead of just the current instruction.  To combat this, there is a setting called "Max utterance per request".  This setting allows you to choose how many utterances the SoundAbout TTS engine will synthesize to your chosen audio output at one time.