Modelling Many to Many Relationship and Counting Unique and Non-Unique Relationships

I have 2 fact tables that consist of unique values that are related to one another through an additional item-relationship:
Fact 1

Fact 2

Item Relation Table - As seen here, not every item in Fact 1 has a relationship to an item in Fact 2, and vice-versa

Currently for visualization purposes, I have the following model. A visual table containing Items A needs to be able to filter another table with Items B down to the items related on selected Item A. (Ex. "1" is selected in one table, and the other table will filter down to A, B, and C)

I have 2 questions regarding this model.

1) Obviously setting up the model like this is going to cause severe performance issues. Is there an alternate way to model this without having to combine Fact 1 and Fact 2 and grow the table exponentially?

2) I need to create DAX measures to indicate for every Item B, the count of times it is the only related item to Items A, and the count when it is not the only item related to Item A. Example below of what the table should look like. Also, preferably, when selecting an Item B, another table will filter down to the related Items A.

Any help is much appreciated!