Increase file disk size without having to grow the file
Since the IOPS and MB/s performance is tied to the size of the disk (P10 128gb, 500 IOPS, 100 MB/s etc.), we should be allowed to assign larger disks to a file without having to grow the file first (as this activity evidently costs more money, but more importantly consumes unnecessary instance space). As a P10 disk would natively be assigned to host an application Db of 100gb, if this DB had a decent read write load from both the application and downstream reporting activities batching data into a DW etc., is not sufficient (not even close..). Therefore you are required to grow the file to 1030gb (as an example) to get this Db onto a disk that can handle this load with ease (the sort of performance customers expect when they are moving to the cloud). But by doing this, you get charged more, but more importantly you loose precious allocation space for the instance (1tb consumed for a 100gb DB, just to get it onto a disk that can handle the expected workload)
Implementing this change would allow more to be done with the instance, without breaking the 35tb or 8tb hard limits on the underlying storage account and instance size. Overall this would allow customers to capitalize on their investment further by being able to add more DB's on without running low on instance space etc. Whilst also accelerating their cloud journey as this service in particular is focused on accommodating for customers moving to the cloud, who are not ready for SQL DB etc. and DB's along with VM's are seen as a nice stepping stone to the cloud.