Does the compression or the expansion of a simple topology potential magnetic field lead to the development of current sheets?

Janse & Low have most recently addressed the following question. Consider a cylindrical domain containing a simple topology potential magnetic field threading its lower and upper horizontal faces, and a perfectly conducting plasma. Suppose that this domain is made to slowly contract or expand in the vertical direction, so driving the field into a quasi-static evolution through a series of force-free configurations. Then are these configurations smooth, or do they contain current sheets (CSs)? We reexamine here their three-step argument leading to the conclusion that CSs form most generally. We prove analytically that the field has to evolve through "topologically untwisted" and "nonpotential" configurations, thus confirming the first two steps. However, we find the third step - leading to the conclusion that a smooth untwisted force-free field is necessarily potential - to be very disputable.