I'm looking to get better usage % numbers from my datastores when it comes to IOPS. I know the default capacity number is 5000 per data store and while that can be changed it still doesn't give me accurate numbers. Most of our datastores are backed by a pool of disks where the data can auto-tier. In the case where you have a lot of low use VMs that move to the lowest tier your potential max IOPS for that LUN has decreased if the data is now on 7k disks vs 15k disks.
I believe I saw this functionality in vCOPS but it's been a while since I tested it. They would calculate an estimated IOPS max per LUN based on the highest number of concurrent IO to that LUN giving you an estimated capacity. This way they don't have to create a storage plugin, while more accurate, for each vendor out there.