This way you just publish a MergeTree table to S3 bucket or a set of static files on an HTTP server, and then plugging it on an empty ClickHouse server.
ClickHouse also wins in the comparison of serverless engines: https://datastudio.google.com/u/0/reporting/c870737c-e8b6-46...
Discussed here: https://github.com/ClickHouse/ClickHouse/issues/43589#issuec...