Creating new data model becomes unusuable after backing out of a new data model read

[FlaggAC]

Backing out of the new data model process is expected user behavior, given the difficulties of getting a good reading over a 30 second period. Unfortunately, that also means the app is unusable for me, because backing out and then attempting to repeat the process results in the walking phase of the calibration (first part) never completing and never prompting a countdown for the jogging phase. I have to restart steamvr every time it happens.