I really like Flask, but Flask's entire point is that it's a thin, user-friendly wrapper around Werkzeug. Werkzeug isn't my favorite low-level web framework for Python (that'd be WebOb), but the fact that it is trivial to dive into Werkzeug from Flask is Flask's killer feature: its whole raison d'être is that you can smoothly and cleanly replace Flask with your own application-specific Werkzeug middleware as you grow big.
So I guess the only thing that surprises me somewhat here is that the author has approached this as a way of scaling Flask up, rather than providing a clean way of moving Flask out in a way that eases transition to pure Werkzeug. I know that Flask has gradually been growing in some ways via its extensions, and maybe that is indeed its future. But I still think that Flask is best used as the training wheels to get a small app up quickly, and that gradually removing it as you grow is likely the best course of action. Otherwise, I think you'd likely be much better off using a framework like Django, with all the bells and whistles you need for any normal CRUD app, rather than rolling it all yourself in Flask in the first place.