Maybe to clarify some things, I’m using the alexa-remote2 package (same as the other Alexa app) for all interactions with the alexa api, so if a feature is not supported by this package then it won’t be supported by my app. The difference between the other Alexa app is that as far as I remember it uses a two year old version of the package so there might be a bunch stability improvements missing and that I use a more generic approach how to support devices.
I need to check the package again but as far as i remember you always need to run every command/api call against a device or group. I might look into supporting speaker groups which could maybe enable announcements across multiple speakers, but in the meantime you either need to send a command to each device or call a routine which does it on Amazon’s side of things.