I think this is a great idea - I think you just got slightly the wrong end of my argument.
I was talking about pain and what you measure/benchmark the app against being subjective.
I know the App is meant not to be subjective, but you have to measure it against something - with a lung infection you can check if it is there with 99% certainty after ResApp says it is.
With pain, all you can do to validate the App is benchmark it against another subjective test, such as observing the patient to give an Abbey score. This results in saying the App is within x range of the Abbey, but not 'better' than the Abbey.
I'm just saying it is harder to do than an infection, not impossible, as you can use multiple pain assessment methods and things like scans to watch the pain centres in the brain fire...
So to summarise, FDA approval might be slightly harder, longer or more costly than ResApp. But this is one negative I see in a sea of positives.