The reason a Star Schema exists is because it reduces the amount of data stored and it reduces the load times (faster updates) because only the facts and measures get updated. There is also little redundancy in Master Data (dimensions).

in the world of fixed assets (on-premise data warehouses) this was a concern, I'm curious to see how this plays out in these Cloud Providers because to me it seems like they will be more than happy to rent you as much space as you want and everyone is happy until the bill comes due.

Doing materialized, flat tables everywhere is great for reporting performance but the tables will not be updated as quickly, there will be redundancy in storage and it will be difficult to sync time dependent dimensions.