If anything wouldn't this argument suggest you should isolate and quantify the active ingredient? Certainly there is more variation between two espressos than, say, two 80 mg caffeine capsules.