Toyota would like to use Cluster Level change streams to write out data to topics named almost the same in Kafka. They have almost 1000 collections that they would need to fan out to.  Today it is limited to 100.