I have been exploring Analytics Hub exchanges and Data clean rooms in google big query. From the docs:

BigQuery data clean rooms are built on the Analytics Hub platform. While standard Analytics Hub data exchanges provide a way to share data across organizational boundaries at scale, data clean rooms help you address sensitive and protected data-sharing use cases. Data clean rooms provide additional security controls to help protect the underlying data and enforce analysis rules that the data owner defines.

After playing around with data clean rooms, I found the below security controls that data clean rooms provide:

1. Can select all vs. a subset of columns
2. Can specify “Analysis rules” such as Aggregation, Differential privacy or List overlap.

The “Data egress controls” are present while creating an exchange as well.

Now, these analysis rules can be enforced on views outside of a data clean room (in fact, some articles I read had the same workflow, i.e., creating a view with some analysis rule and then adding the same as a listing in a data clean room). Since once can share the same view (with the analysis rules enforced) through an analytics hub exchange, I see the following difference between analytics hub exchanges and data clean rooms:

1. The unit of data sharing (listing) in an exchange is a BQ dataset whereas in the case of a data clean room it is a table/view.
2. In case of an analytics hub exchange, the linked dataset tables/views refer directly to the tables/views in the source dataset whereas in the case of a data clean room, a new view is published in the source dataset which is what the linked dataset tables/views refer to (exception being a view with analysis rules in which case no new view is published in the source)

I expect that in both cases, it is possible to add listings across projects/orgs with the right permissions granted.

So, one can always create views with the appropriate analysis rules enforced, collect them all in an authorized dataset and then share that in an analytics hub exchange. Additionally, in a data clean room, it is not possible to use an arbitrary query to specify the view or to even change column names. So, if the use case involved anything other than column selection or enforcing the analysis rules, such as renaming a column, one would first have to create a view and share that view as a listing in a clean room, and since the view does not have any analysis rule enforced, the data clean room will create a new view in the source dataset essentially leading to two duplicate views. Also, a clean room automatically creates a view to avoid direct access to source tables, which seems more of a disadvantage rather than an advantage because there is no control over where the view is created, and one can also always create the view first and share it in an exchange.

So what I want to understand, in light of all the above, is what do data clean rooms help accomplish that an analytics hub exchange cannot already do?

1. One difference is the way the data is organized as I mentioned earlier (dataset level for exchanges and table/view level for clean rooms). Is this the main reason for data clean rooms, that in case of multiple parties, everyone can add listings to the same clean room and then import the clean room as a single dataset, whereas in an exchange, all parties would need to create separate datasets as listings and then import them all.
2. Secondly, I guess the data privacy restrictions would be visible to all parties in the case of a clean room whereas they would be hidden in an exchange.

Are these the main benefits of a clean room?