These envvars are only used when we know we can depend on the Database ID to be unique (when there's no chance the database has been cloned). Requiring certain levels of entropy in these envvars thus doesn't add much -- the true uniqueness will always be coming from the auto-generated (random) Database ID.