Title: Expose metrics such as latency & rps by VHost

Description:
I am looking for a way to visualize metrics by VHost or Route. Currently, they are exposed per cluster. I did a quick ask in Slack and found that other users are parsing through the access logs and deriving this information from them. I'd prefer to see if there's a way to have them come through the included metrics mechanism that exists today.

I'm curious if anyone has similar needs in their environments, what places in the code would need to be touched, and if I could help possibly implement a PR.