VNX Monitoring and Reporting - Performance Report re IOPs @ latency

I am new to VNX and just installed VNX M&R recently. The task I need to accomplish is this: management wants to know when the VNX5300 we have is starting to run into small performance bottlenecks (before it gets bad). They want a weekly report detailing some metrics that are deemed most useful for determining the "rising water" level of increasing IOPS or latency on disks/LUNs as we put more and more load on the 5300.

I know enough that a primary indicator of load is IOPS at some level of latency deemed too high (5ms on write, 20ms on read, usually) but is anyone familar enough with VNX M&R yet to know what counters I should be looking at? The counters on a LUN performance summary report for example have "response time" which is described as "The average time, in milliseconds, that a request to this LUN is outstanding, including its waiting time". This seems like the correct counter but is it accurate? The tooltip on the column says it will go status "major" at 20ms and critical at 25ms. But is this the correct benchmarks that I should be watching for? What about queue length or something else?

Thanks for any thoughts. I know there's no silver bullet answer.

Pete