We already know, for instance, a bit about the beginnings of glucose dysregulation in the beginning of diabetes. If we could have more samples of "continous" glucose data from a person, then we could catch glucose dysregulation before it becomes either type 2 or type 1 diabetes (or something else). We'd thus be able to more easily keep someone at risk from getting type 2 diabetes through simple lifestyle interventions (which would be a lot easier to study with a lot of continuous glucose data) or if it is type 1 we could administer the BCG vaccine to keep them from losing their pancreatic function. That sounds like a revolution to me!

Also, for obesity, it is known that lactate (and other metabolite) levels predict someones metabolic health and whether they have metabolic syndrome. If we had continuous lactate data or maybe BCAA data (this is probably one I'm not sure they can get easily with current tech) then we could find out when and why metabolic syndrome happens and how best to stop it and reverse it.