For your discovery of the workings of this theory I think it's important to note that after
the magnets are moved into position a field is established. The field does not interact smoothly but it establishes after
the magnets got their position. And while
the attractive or repulsive magnetic interaction does a pulse of work A new set of values is inducted in both magnets.
For us this means we can move a magnet towards another then have both interact for a moment and move it out of the field before this interaction goes "out of hand". Thus therefore it is important to note we can speed up the apparatus so-that we only get 1 moment of interaction.
If a magnet on a wheel approaches 2 other magnets of which 1 repulsive and 1 attractive it should be possible to tune it in order to present no drag on the wheel at all. If the wheel now spins fast enough so that there is only ONE interaction we can get repulsive and attractive work out of our 2 magnets.
But this work is only 100% free for 1 moment. Obviously, there will be more attractive as repulsive force after the repulsive magnet is pushed away and the attractive magnet is attracted. Our next instance of magnetic build-up will not be free of friction. We can never get it to speed up all the way to a single viscosity moment. The fun part is that the remaining repulsion is still subtracted from the increased attraction. The moments just before and just after the mini moment of free gain are also overunity. Here is a simple illustration. (actually 2 of em)
In the top image
the magnet at the left can move free towards and away from the 2 other magnets. The bottom image
is the situation after doing work on the magnets at the right. The gain is 2 times 100% over a distance. But it only takes 20% to remove the magnet again. This would make the interaction 580% efficient. gn0sis.com- Energy from magnets