That's exactly the point (ie just prior to distribution) where a simple sanity check should have been run and the config replacement/update pipeline stopped on failure. When they introduced the 200 entry limit memory optimised feature loader it should have been a no-brainer to insert that sanity check in the config production pipeline.