http://www.cheniere.org/correspondence/031808.htm
http://www.cheniere.org/correspondence/042208.htm

has anyone here tried making an over unity device with asymmetric magnets?

But now suppose we contract with some nanocrystalline folks, to build a special "laterally asymmetrical" permanent magnet.

Nothing special needs to be made to do this, just build a Halbach array using common block magnets.

Suppose we mount the rotor (say, S-pole facing the stator) and stator (say, N-pole facing the rotor) so that, when the rotor S pole is rotating and approaching the stator N pole, the strong sides of both bar magnets are facing each other. This gives a certain acceleration added to the flywheel and shaft, and it stores up a certain amount of free angular momentum in the rotor and flywheel, while the rotor is traversing through that ?forward mmf? zone.

This rotor would still not spin because as the magnets pass into the "weaker side" they will try to return to the stronger side and thus the extra energy will be lost. Tom needs to play a little bit with magnets before he blabbers stuff like this.  Sure it sounds good but does not work in real life. 

Tom also doesn't explain back emf very well either.  If you energize a coil with a DC battery, current will be flowing in that coil at some constant rate.  If you move a permanent magnet toward or away from that coil, even while it is energized, the magnet will induce a second current within the coil.  This second current can work for or against the energizing current.  If you moved the magnet such that it repelled against the energized coil, the induced current will add with the battery current.  If you do the opposite, so that the two attract each other, the induced current will cancel the battery's current. 

The induced current does not cooperate!  It will make the repulsion force between the two stronger (so that trying to push them together becomes more difficult) and it neutralizes the force when they are attracting (bringing the current in the coil to near zero as it attractively approaches).  This effect is most easily seen in motors.  When the motor has no load, the current flowing in the motor is almost zero because the magnetic field in the rotor is neutralizing the current in the stator coils.  When you apply a load, it slows down the rotor, the input current to the motor is no longer being neutralized by the rotor's magnetic field and current starts to flow in the stator windings.  This flow of current increases the magnetic field in the stator and forces the motor to continue to spin while under load. 

To me, this back emf is an unbalanced force.  There is nothing in the system acting to cancel it out.  Unlike friction, back emf is only harmful to the system when it is allowed to generate a magnetic field.  Where as friction is there all the time.  I feel this should separate it from a friction-like force (even though many define it as such).  Modern electric machines allow the direct interaction between back emf and the prime mover.  Therefore, they allow this unbalancing force to perpetuate itself, completely rendering the machine a simple energy converter instead of energy creator. 

If a machine were designed where induction could take place yet the counter force from the coil could not interact with that induction, I believe we would have the makings of a balanced system.  The geometry of the machine could then be employed in a way which all negative forces canceled each other out, allowing easy rotation of the prime mover with no counter forces other than friction. 

Haha, now the question is, how do you build something like that? 

Beats me!
Charlie