Wdnering - This is exactly what I was trying to say (you said it better). MNKD will get early reads on how large homogenous populations performed, but the devil is in the details. What if afrezza works great for the general population but not so well for participants that fit a subset population?