If you cancel winding inductance you get nothing. LEAKAGE inductance is what you are worrying about and is a different animal.

Your bifilar coil if connected as two windings will produce power but will do nothing for leakage inductance, that is related to slot geometry, flux path length, flux leakage paths and armature reaction.

If you connect your bifilar windings series opposing it will be non inductive and you get nothing out ( you will at least avoid the effects of the leakage inductance with no output).

I asked questions about the issue of inductance and learned that it's much more complicated than I previously thought:

http://www.fieldlines.com/story/2006/7/4/425/44513

I obviously wasn't asking about the same thing, but the response was so informative, that I keep that thread printed out and in a binder with other "useful thoughts" that I've picked up along the way.  

If you were to draw out a diagram of one phase of, say, a 4-pole stator, you would have a pattern like a four-leaf clover.  If you wound a phase with two separate lengths of wire, one going clockwise, the other counter-clockwise, then all that would achieve is current flowing clockwise in one, and counter-clockwise in the other.  The reversal of the winding direction and the reversal of the current direction makes two negatives multiply, leaving you with the same positive result.

lot of good answers, some close to what i am trying to ask, others while
informative are not quite what i am looking for.

yes i understand that we have all sorts of stuff going on
IR losses, iron losses, hysterisis, eddy currents, at least 5 different forms of reactances, armature reaction, leakage inductance, slot leakage, and all sorts of other things.

some designs are dominated by part of the above list, some others dominated by others.

after reading some of the responses

how about this example

a single pole standing on the page vertically, lets make it a nail in a board

onto this nail we wind a coil clockwise and bring its leads out to the left of the page

then we wind another coil counterclockwise (equal turns) and bring its leads out the right side of the page.

there will be no interconnection of the coils

we now pass a magnet rotor over the nail head,

we get induction into both coils, as the magnets pass we get a positive pulse in one coil and a negative in the other when the north magnet pole goes over and the opposing effect when the south magnet goes over.

so far so good.

now we apply loading to both coils, seperate but equal loading.

the current flowing through the coils will affect the flux at the airgap
armature reaction, and will increase with current in the coil, but

with opposing coils the back induction that rises with current is offset by the opposing coils so therefore there is no armature reaction (in theory)

the question that remains is

would the back induction of one coil try to fight and reduce the output of the other coil and visa versa?

(i gotta learn how to make a sketch and post it)

the reason i ask is based on the fact that some alternators (large ones) use a second winding afixed to the top of the pole pieces that is shorted or shunted like a shaded pole in order to reduce the armature reaction, and
i am working on trying to figure a way of reducing pole reactance, and reducing pole inductance if possible is a step in this direction.

anyway sorry if i try your patience guys
i have only spent about the last year (intensively) studying alternator design
devoteing about 2 hours per day on average to reading everything i can find on the subject.
perhaps in a couple years i will know a bit more, but i suspect i will only have more complex questions.

It makes a difference if you connect it 3 phase or 6 phase if you rectify it but that is a peculiarity of the rectifier.

The damping cage of a single phase machine primarily prevents the pulsating mmf penetrating the field circuit. For polyphase windings it damps pole swing in paralleled machines and also provides a means of starting synchronous motors.

Not sure what you are trying to do, it must surely be with iron poled machines with small air gaps. Most of these snags vanish with neo with effectively very large air gaps and high coercivity.

Winding 2 coils per pole is not going to solve anything and mutual inductance is a complete red herring.

The air gap is often a compromise between magnetizing force, tooth tip losses, windage losses and cooling requirements.
Older texts talk about this extensively, though often in confusing terms like armature reaction. (none of this is complete in any means).

Basically your desire to reduce losses is in a sense like asking how do you design a car.  Per unit losses as a function of cost is actually a sh*t ton of variables.

time to strip the core, and start the rewind

not sure whether the bifilar approach is worth the effort because the current levels will be so much lower than the original design anyway.

dramatic reduction in current should provide a substantial reduction in air gap leakage or flux shift anyway,, not sure if the counterwound bifilar would provide a measureable gain in efficiency even if it did work.

thanks again