What is the usage of the Flow database that is required in the installation

We have a running instance of Flow on a Redhat server and a remote postgres db.

My question here is:

How or in which cases Flow uses its own database,

I have Flow connected to the foundation database to do some queries.

But it seems that is not using his own database to store or do high loads,

I want to know if it makes no difference to have the Flow database installed on the same local server as Flow.

I have read the documentation and it mentions the local installation as the default option.