Thanks for the clarification :
When possible, mute Homey to prevent double voice messages
This is possible with flows or just mute it using the app.
I would prefer to send voice commands with having to use the additional “Ask Homey”
This is no longer necessary, there is full google integration in homey now. You can read about it here : https://blog.athom.com/homey-now-works-with-google-assistant/
I never used a HTTP request. This is something I want to look into
You can read about it here : https://github.com/athombv/homey/issues/2666
Would this support the ‘ask for confirmation’ part?
As far as I am aware this is only available using the alexa app, and even then its note a great experience. (alexa asks for confirmation, you say ‘yes’ and while that “works” alexa will respond with a “I dont know what your talking about”… I’ve found it more useful so that Homey can listen for when you ask Alexa for the tv to be turned on, or play music etc, the homey sets the lighting.
I am not yet familiar with the last two options you mentioned. I never used a HTTP request.
It requires that you have the relay running locally https://github.com/greghesp/assistant-relay - and the url request app from the app store to ask it.
The TTS app works, however there is a lot of latency and the voice is very slow.
If youa re only using the same messges over and over, perhaps you can instead use the soundboard app locally to supply the files for sonos to play? https://apps.athom.com/app/com.athom.soundboard
From what I understand you would like to have two sonos devices and (now) google device, and want TTS . There are five methods to achieve this with Homey. While I understand you are restricted due to features not released with in the Netherlands and being unhappy with latency (when making pretty intense requests) - I personally can not think of a way around these restrictions that could be solved with a sixth Homey app.
Best of luck, if you do find something which meets your criteria please be sure to let everyone else know