We had one client who had Time as part of the code in a list, which is essentially what you are doing, and the list was 130M members. Once they took time out and dimensionalized it in a module, the list was only 6M members, which also brought the overall size of the model way down.
It is apparently the same but it is also dependent on the combination.
In view A, your product and location are repeating for same time and amount. if you read by yourself then you should read the product 3 times in view A.
In view B, you should read the product only 1 time, but total combination for both view A and B are same, but there is a difference you are making Anaplan to read the product list only 2 times in view B but for View A Anaplan is reading it 6 times.
So if View B is better than View A in total import time, then it should mean that Anaplan performs with reading less list items than line items. Which is similar to Transactional list analogy of removing time.