It is my belief that a magnetic field is effectively a moving mass, involving constant rotation of the ether from one pole to the other.
So, these are my thoughts and propositions...
In its natural (complete) state the magnet exhibits constant field motion with the most intense point of field focus at its ends. The centre of the magnet however can be observed as at equilibrium or, in other words, of zero field intensity in exact middle of the unbroken rod.
Unlike breaking two separate magnets apart, breaking a single magnet however has about the most violent effect on the molecular bonds (and associated fields) that you could induce. At the instant of the break there is a very sudden and an immediate imbalance in each of the two pieces of the magnet. As the square law applies to distance and intensity, it can be appreciated just how violent an effect this would have when imbalancing the field in a magnet at molecular level.
As the field is most concentrated along the axis of the core material, the 'flux' within the core itself will immediately try to regain equilibrium within each piece - effectively creating a "rush" of flux from what was the outer pole to what was the null zone of the unbroken magnet.
The net effect therefore being to switch the polarity as, where before there was an observable EXTERNAL North-South flow, there is now an INTERNAL North-Null flow, overcoming the inertia of ether spin and thereby reversing the external (observable) field.
I'm not at-all confident that this effect could be induced by pulsed coils and such as the speed by which a coil might be able to distort the natural flow within a magnet would be many orders of magnitude less than the speed at which molecular bonds are broken when the magnet is physically snapped. I could be wrong of-course as this is conjecture in uncharted territory, though an applied field of sufficient intensity may simply mean the the magnet is re-magnetised in the traditional sense.
It is also my guess that the magnet polarity will not flip if gradually cut through as it would be possible for the magnetic field to equalise gradually - with rotational inertia maintained by ether spin throughout. However, it is also my guess that if the magnet were cut to something less than half its full (magnetic) width and then snapped, only then might there be sufficient field imbalance along the centre axis to overcome rotational ether inertia and bring about a switch in polarity in each piece.
I'll be watching this topic for further experimental results to see if my proposal is supported by experimentation.