Why should I add a pivot to the data model when the underlying data is already included?

If I create a pivot table based on a data set that is in the data model, it gives me the option to add the pivot table to the data model. What is the reason to do this if the data that the pivot is based on is already in the model? Doing so seems to create a copy of the underlying query.

Thanks,

Ben