It's good we can agree on this aspect.
It all comes down to number of legs in the system. 12 full coils will have twice the EMF per rev as a 6 full or 12leg serpentine winding will have. How we arrange the coil ends is of no consequence to the EMF generated... simply put 12 coils will have 24 legs to cut, and 6 coils only 12.... twice as many same turn coils.... twice the EMF.
Stacking factor and physical constraints and resistance are a different matter. I suspect if you try hard enough, the outcome should be very similar, and stacking technique will be the difference in the resistance. The coils ends will be the same for the same turn number, only the coil interconnects of the full coil design should be the difference. I understand your experience was slightly (8%) different, but if there was to be always an 8% advantage to serpentine winding, then there would be no full coil devices built.
I can promise you that when I rewound the 90 coil AWP stator, if I could have gotten away with half the turns or half the coils for the same kind of result I would have taken it.... sadly this is not so, and magic is only possible when trying to envisage primal forces and planck time scales. (whats gravity I ask.....)
My first guess at why car alternators always have skein windings and almost nothing else does, has more to do with the almost unlimited rpm available, and cooling and space.
As I mentioned earlier, armature reactance is the major factor in auto alt output, and can be mitigated by less turns at more amps (ampere turns.... which would need more revs to compensate, but thats of almost no consequence for car and truck alts. One minute they are running a few thousand at idle, and next can be 18000 at high revs. Volts per turn are not critical. They just need to design for cut in around idle X step up ratio. They also need to design in enough reactance so that things don't go pear shaped when the revs come on,and then we need to self limit. This allows for simple regulators, and no shredded belts...... Imagine if you developed a few hp at idle, what would you develop at full revs if the regulator was turned on full, and the battery and load could take it... even for very short bursts if it didn't self current limit with reactance...... scary.
So thick wire, open coil structure, no care of rpm, ease of manufacture, easy cooling, and current limiting...... skein winding is a good option.
If it offered any noticeable improvement over normal full coil winding, the alternator at the power station would use it.
8% for 24/7 is a massive diesel saving..... but it's just not there.... and I know as I have one here.
Even the exciter stator is wound normally...... there is a reason for normal apparently..
You know it's big when I need to use my trac 955 to unload it.
......................oztules