Using DuckDB WASM and Cloudflare R2 to host and query big data (for almost free) (andrewpwheeler.com)

I didn’t think it was possible to host such amount of data for free. I have not tried duck db but this gives me the reason

Does that approach really work for a 72GB dataset? I assume that means DuckDB will need to load all that data in the browser?

It would be nice if R2 supported Requester Pays like S3. In the past there's various data/files I've wanted to make public but not at my own expense.