When looking at the hardware, HASS OS recognizes and assigns details to the sound subsystem (/dev/snd/).
Why is the snd subsystem not exposed to HASS?
ALSA has been part of the linux kernel since the late '90s.
The library playsound, pyaudio, or simpleaudio are available since 2016, 2006, 2015 respectively (MIT license). There are others with different licensing. Both WAV and MP3 can be passed to the snd system, and audio written from them.
Plug a 3.5mm-t-BT/BLE bridge into the port, and we have the audio solution solved. They range are very inexpensive (search for 3.5mm to bluetooth transceiver)
This works on regular RasPis without any issue. Of course, also works on regular PCs as long as they have 3.5mm jacks.
My primary use for this is not audiophile listening to Vivaldi’s quattro stagioni. It is for assistance, speech and audio commands.
(I am open to be told I do not know what I am talking about. 
)
Alternatively, what would it take to implement such add-on?