With axial alternators with modest size discs there is not a lot of space at the centre that you can really use when you allow for the stator requirements and the hub and fixings. The power you could get from the small space available would be quite limited.

It is true that in low winds you can get little anyway so if you use the centre bit for low wind operation it may work out reasonably well. It is also likely to be the case that using the small space available more effectively with slightly bigger magnets on the main winding would be more effective.

With all these 2 machine schemes you have the same problems as using two windings on one machine and the 2 winding scheme probably wins out. The low wind machine or winding will become inefficient in higher wind speeds and you will need to cut it out of circuit or tolerate dreadful losses if it doesn't burn out.

It can be done, it has been done and in fact I have tried it more than once. It comes into a similar category to star/delta or series parallel switching but has none of the advantages of either of them. With star/delta or series parallel you actually retain all the winding and rearrange it to suit your needs rather than have only a part winding under both conditions with your approach.

I am sure that there is a lot more that could be done with all these approaches and I find it strange that people cling to the one winding approach at all costs, it most likely comes down to simplicity in the end. All switching schemes involve some degree of complexity, even star delta is far from simple and few seem to have attempted it.

I got part way along this road several times and had quite reasonably working schemes but as you refine the switching arrangements you eventually reach a complexity that is not justified for the gain when less complicated switching in the form of a converter lends an infinitely better solution.

If you want to try I suggest you use bigger magnets rather than try the second alternator and use 2 windings on the main alternator. The low wind part need only take about 25% of the space. Use it for winds up to about 12 mph then ditch it and let the main winding cut in. Even with 25% of space wasted for the main winding the rise in cut in speed lets you use less turns of thicker wire and you still end up with a more efficient alternator capable of a much better match to the prop in higher winds. If you can devise a simple mosfet switch to disconnect the low speed winding then you are well on the way. If you get that far then you could consider keeping the low speed winding and use the mosfet as a buck converter to alter the voltage range and keep the winding matched into the higher wind speed range.

This could be quite good approach and I have done it on a test basis. If you can manage to keep the line resistance to the main alternator negligibly low it is an excellent scheme with a smaller and simpler buck converter than required for the whole machine.

If you use suitable fuses to isolate the converter if it fails it has the considerable merit that the machine will still perform satisfactorily well in most winds until you can deal with the converter.

I think it is all in my diary entry about "matching the load", probably I called it hybrid schemes.

I hope you can make something of the first bit of this reply even f you are lost on the last bit.