Hello! I'm a database focused Field Solutions Architect... specifically I've been working with SQL Server as a DBA/DB Dev for the entirety of my career.
I would encourage you to consider measuring and evaluating your workload from a different perspective, not just purely IOPS. The thing about IOPS is that it is meaningless unless you also consider average I/O size. 1 million IOPS where each I/O is 4KB will be far different than 1 million IOPS where each I/O is 10MB in size. Think of it like, I can move 10 shoeboxes quickly from one room to another but would take much longer to move 10 large suitcases. 10 "containers" -> 10 IOPS, but each of differing size.
Also, when working with customers, I find that more often than not, bottlenecks manifest elsewhere in the stack, whether it be in your storage fabric, a hypervisor, or even host CPU hitting a ceiling and thus not being capable of driving more I/O.
For your specific situation, I would suggest engaging your account team and requesting a meeting to review performance of your current arrays. They can work with you and if necessary, bring in a Database FSA like myself or one of my teammates.