I'm thinking that now with all these agent to agent frameworks this could potentially work for that. If you send off one agent you want them to keep up to date and sync / talk to each-other. Could your solution work for something like A2A by google or similar to enhance the synchronization across the different agents doing their tasks and prevent them from landing in a loop or similar.

If the consumers stall, doesn't the WAL have to grow in unbounded fashion? Does it place any backpressure on the writers?