Improve performance for very large postgres schemas

Hello! We are using Debezium to track changes to a single table in Postgres and have an extremely large database (~350 schemas, ~200 tables per schema, ~15 columns per table). When Debezium starts up it seems to be loading the entire schema, which for one of our production-like environments can take up to a half hour of constant querying to load the required data. Since we really only want the events from a single table, is there a way to limit how much of the schema that Debezium needs to load? Or is there a better approach to this problem?
 
We have snapshotting set to `never` and are using the table whitelist.
 
This was originally reported on Gitter, and I wish there was a way to link to a thread so that I could attach it here, instead I've copied it into the comments below. I was advised to convert the thread into a feature enhancement.