That’s the use case for cdc, to make it equally easy to use a DW. As always the complexity is just air you move in the balloon. The oltp db can spit out the events and forget them, how you load them efficiently is now a data engineer’s problem to solve ( if it was easy to write event grain on an olap you would not need an oltp). Kafka usually enters the room at this stage and the simplification promise is becoming tenuous.

that works great if all the data you need for reporting is in the database you're replicating.

You'd likely want a data warehouse if you also need to report on data that isn't in your prod database (e.g. stripe, your CRM, marketing data, etc.).

If setting up a data warehouse, etl, BI, etc. sounds like a lot of work to get reports, you're right, it is.

shameless plug: we're making this all much simpler at https://www.definite.app/