You can make magnets going side to side do something, it just hasn't been demonstrated it is more efficient use of the magnets. Even in the most optimal situation, in a Halbach array, it takes three magnets to get twice the output. So the improvement just doesn't scale up to an advantage. The effect also doesn't work well on a smaller scale because the like poles face each other at greater angles relative to a decrease in pairs. As you move away from the perfect perpendicular angle, the power output drops off. But there are a lot of ways to skin a cat. Maybe you are tackling the issues in a new method.
Nobody is insinuating that you're not smart. I think people are just struggling to see your premise. So we wait and watch.
good explanation, but it actually takes 4 times as much magnet volume with an ideal halbach array in order to get twice the flux density in a given volume. this is because the energy stored in a magnetic field is proportional to the square of the flux value, multiplied by the volume.
so anyhow, this "new" method of using magnets oriented sideways to push the flux out the sides of the disk, only works if you have metal pole pieces between the magnets in order to conduct the flux out through the disk so there is no wasted air gap. and you need coils on both sides so you don't have wasted leakage flux on the other side of the magnet disk. the only reason i can imagine doing it this way is if you have a very high speed machine and want to use laminated pole pieces, and some kind of carbon fiber disk to keep it all from exploding.
without the metal, it is no different than locating the magnet further away from the coils.
some improvement could theoretically be made for a traditional axial flux machine if you use sideways orientated magnets to push more flux through the axially facing magnets. but, with the backing plate shorting out part of the magnet you added, there is little point to do this.