Why does feature do redundant sending

When a user does a new facial expression, it is sent from him to all other users via the video bridge. But I don’t seem to find any handling of received facial expressions that got sent that way.

The same expression update is also sent via the xxmp speaker stats module and received expressions are handled that way too.

Why use the video bridge channel at all, when:

1: only sending is handled but no receiving?
2: The emotion recognition feature already works perfectly fine because it also uses the xxmp server/speaker stats module to communicate the same updates?

The videobridge broadcasts seem to be redundant and therefore unnecessary, or am I missing something

Facial expressions is something new we are working on.

We do not want to load prosody with many messages that’s why we use the bridge channel for live updates from client to the other clients.

The speaker stats is used for history, so when you connect later in the meeting you will get the history, and one more thing, that at the end of the meeting you can get an artefact with history for the meeting.
The same happens with dominant speaker, the only difference is that expressions are produced from clients and send to bridge, where a dominant speaker is produced from the bridge. But both are coming over the bridge channel to clients.