This watch can be considered as a combination of 2 independent twin movement which were born under a hight accurate adjustment (~same DNA). In it, there is a pinion to let them communicate to each other. Let think them as 2 swings in playground but they were connected by some way(the rack and pinion in the center of the movement). What watchmaker wanna do is make them beat in a same beat (it is the easiest way to achieve resonace), so same inertia and length of 2 swings are needed to be the same under gravity. Then watchmaker worked hard on the train design and finishing to make sure the same pushing force on these "swings".
If all are perfectly worked out in these mircoscopic "swings" on your wirst, then we do not need a resonace watch, Just because nothing can be perfect, what we are doing is getting closer to perfect. That's what a resonace watch try to beat the error (even twins can't be 100% identical). The beating difference come from a shock, a tiny dust which stick on one of the thousands tooths or simply the systematic error. So they may beat in different freqeuncy, and result in difference on the time of 2 dial. And the aim of this watch is to get the mean beat rate of 2 escapments by resonance, but not simply read the time by observing the mid point of 2 movt. But I think it is still meaningful to read the mid point which are really compensate the error happened on 2 movt. By some kind of source, this error tolerace is 5sec/day. And some user told they were not synchronising after certain time. Moreover, FP Journe put a second zero setting mechanism in this watch to let them synchronized again. Some watchmaker may considering to show us the mid time but not 2 times in future?