Sure, we should not fork if it is possible to develop the functionalities and test them within the current framework.

I just thought it would be easier to concentrate on the quiz part, working with REST APIs and a JS frontend.  Because the app will be offline, the normal quiz engine for validation and saving responses may not work.  I may be wrong.

Can you please comment on the desirability of the idea proposed above? 

Also useful would be broad directions on where to start, please.  So we can plan and commit resources and timelines.