I agree that's ideal but probably wouldn't expect a latency optimization expert to understand my business well enough to do that and that's fine. But I -would- expect them to say something on the level of: "this is the kinda latency that will prevent you from ever moving to a cloud desktop" or "at this level of latency, you are going to suffer X second delay every time you interact with a complex web app like Jira" or "you are an HFT firm and I know that your competitors are at about 1% of your latency." Then let your client decide if this is something that matters to them. Quite different than "wow the CTO is too dumb to understand my genius"