RLS for 2 DIM Tables (Combinations)

Hello PBI Users and Experts :),

I am recently struggling with an easy concept of RLS for 2 Dim Tables.

The thing is I want to make combinations of RLS depending who will see a report. 

First I will share a simple data model to ilustrate the problem.


And a table with needed result for RLS:

So the thing is I have a person that wants to see sales for ALL Segments for ALL Departments except Pharmancies, in which this person should be able to see sales of only Marketing.

I wish if anyone can explain step by step how to do this. I looked in so many places for solutions to this concept but to no avail or I could not understand/comprehend.

Of course I could make combination in Sales Table by Concatenate() foreign keys and put RLS in Sales - But to my understanding it is not recommended to install RLS on Fact table for performance issues or data flow.

There is also second option for Intersect() DIM tables and connect big DIM with Fact and install RLS in big Dim Table - But I do not think it is the way for this case.

My question is if there is any clever way that solves my problem.
Below I'm sending the case file. (Don't know how to Upload File, so below embedded file in OneDrive)
RLS Case - Pbix 

Thanks a lot 🙂