Can it (or could it be made to in future iterations) record the sort of data used to produce the output of an ECG (echocardiogram)? That would be quite the application for people with actual or high risk for hear disease, I think?
No chance you'll detect a QT interval change with this.