Lets be clear about this. Magic is magnetic fields and electric fields and gravity fields...we dont know what they are .... at all..... but apart from that there is no magic. We do know all we need to know about their effects and behaviour.
To induce an emf into a wire, it needs to have to see a changing field, and ideally at right angles to the wire direction (angle of attack for want of a better phrase)
It does not matter what winding style you care to use, if you keep the flux density and rate of change the same in all cases.... then it follows that you MUST have the same turns number for the same emf at the same rpm... no magic can change that.... very simple.
So all the winding styles have to achieve that same thing. The difference between like serpentine and like flat coils in epoxy or resin can be small or considerable.... they are hand made, and usually to accommodate the practicalities of the stator size etc.
If you try to emulate the 4:3 with skein, then the end loops will push the resistance up for the same emf (huge end loops)....... so to get around this folks try the magnet next door , or every magnet or any combination they please, but in the end it is the number of wires getting cut at the same time in the same phase that will generate the result. (angle of attack assumed the same)
If winding with serpentine or wave, then the chances are high that the flux density will have to change to accommodate the crossing of wire in the gap. With steel core this counts for very little. The Alts for cars take advantage of this. Very thick wire, few turns, and the slot directs the field...and the field is is at the behest of the regulator so all problems are solved. Air gap machines don't have this luxury, so for skein, or serpentine or layered, these problems are far more important to overcome, as are the end loop radius.
Serpentine does not solve the end loops. For every 2 bundles of wire in the legs, there are 2 loops to support the current path. eg for 6 poles, a single layer skein will cover three poles. It will encompass 6 end loops for the three poles..... same as using three coils. same result.
If we then inter wind a longer skein to encompass the 6 poles in a single skein, we end up with 12 loop ends, and 6 poles covered.... same thing with flat coils ie 6 coils, and 12 end loops. Serpentine does not get rid of any end loops. It just saves some practical winding problems.
The only thing you can hope to change when using any wiring scheme is the resistance (for the same gap and magnets), but the emf induced in a single turn in like circumstances will be the same regardless.
If somehow you can pack thicker wire of the same turns.... then it is a win. I cant see overlapping doing this, but if you found a difference, then it will likely have to do with flux density being different, or the attack angles on the legs are different, because the same turns will yield the same emf............unless you use geometrically different shaped coils I guess. With iron core that is not possible to mess with, but with free hand air gap..... sky is the limit I suppose.
So if you see an appreciable difference in an experiment, and it does not match the theory, then wee need to look for why. We understand very very well how electrons are influenced by magnetic fields, and this is the basis for tv tubes, and an endless amount of electronics all the way to atomic accelerators.
We can calculate for the tiniest variations... we cannot break the rules. How we interpret the results of experiments needs to be done very carefully, mindful that the rules cant even be bent.... so look for the real reasons something may not follow the rules.
When it come to theories that are not based in true science, then it's just a theory anyway.... otherwise known as an opinion, and society seems to be in love with that "science" at the moment.
................oztules