It has started on it's own in heavy winds, and the blades are out of balance,
That's your problem. Get the blades into static balance and it should start on its own.
Static balance is easy.
one thing I noticed is when I do kick start it, the hole machine then moves a foot counter clock wise, so if it's facing south when I kick start it, it then moves towards the east maybe a foot.
Another sign that it's out of static balance. WAY out. (And that you've got a really good yaw bearing. B-) )
I'm going to cut up a new set ...
Just balance the ones you've got.
As for gap, it should not matter at all (unless it results in the stator and rotor rubbing). An alternator without iron cores (and without magnetic metal screws near the rotor magnets) has no cogging. So the only things keeping it from starting when it's dead-stopped are bearing/seal friction and static imbalance.
I'm going to check the rectifier for a short.
Won't hurt. Two shorted diodes - one to (+) one to (-) - could put current through a coil and hold the prop in a preferred position. Just one will put drag on the mill like a shorted coil for part of the rotation. But drag from a shorted coil or diode doesn't happen until the mill is actually moving.
(Drag might keep it slow enough to exacerbate the startomg problem from the imbalance. On the other hand, enough of an imbalance that "kick-starting" the mill yaws it by a FOOT is a LOT of imbalance. So I'd deal with that first, and just check the diodes for completeness and because it's easy.)
