SQL Server 2019 BDC Datapool

Question

https://learn.microsoft.com/en-us/sql/big-data-cluster/concept-data-pool?view=sql-server-ver15

I am still get confused for the storage on SQL Datapool .
SQL data pools consists of SQL Server instance (2019 ) .

1. Can I create table DIRECTLY in these SQL Datapool? Instead of create a external table from Master instance ?
2. What's the storage of these table in SQL Datapool ?? mdf/ ldf ? and are these files in HDFS ??
3. these table support in-memory ?

"In order to enhanced performance across large data sets, the ingested data is distributed into shards and stored across all SQL Server instances in the pool. Supported distributions methods are round robin and replicated. For read access optimization, a clustered columnstore index is created on each table in each data pool instance."

how many number of distributions are there? In Azure synapse it's 60 ......... And in the case round robin , seems not the best distribution .....as there will be so many data movement between distributions.