You say the magnets are 1" dia but you don't give the length ot the gap between the magnets.

My immediate thought is that 12 magnets on a 14" disc are spread pretty thin for an 8ft prop.

If you give the magnet length I can see if you have any chance of a reasonable match to a 8ft prop.

I suspect that you ought to be thinking of 16 poles with 12 coils or ecen doubling up on the magnets with 2 concentric rows of 12.

You  can manage with less magnet material and get good results in light winds but if you want the full output of an 8ft prop ind winds of 25mph you need a decent amount of magnet material.

I realize your question was addressed to Flux, so I hope you, and he, don't mind this "hijack". I hope he will comment on these suggestions as well, as I value his opinions.

This basic sketch shows eight poles for simplicity, but any even number of poles could be used.

To achieve the thinest possible air-gap, in the actual construction, the coils would be "nested" so that the stator is actually only "one coil thick". In other words, all the overlapping would be done at the rounded ends of the coils, where there is more room.

With regard to coil design, and determining the optimum wire size, the largest that you have space for will have less resistance/losses.

In lieu of a fully engineered design, I suggest the following "trial-and-error" steps.

1. Determine the voltage that you must have for your specific application.
2. Build the rotors and the stator coil supporting structure first.
3. Make and mount one coil, using, let's say #14 wire, with as many turns as you can fit within the physical space that will be available for a single coil.
4. Connect a variable load to the coil, and spin the rotor at the desired "cut-in" speed, and measure the voltage and current.
5. Evaluate the voltage at various loadings to determine if you need more turns of smaller wire to increase the voltage, or less turns of larger wire to increase the current, bearing in mind that you will be connecting the other coils in series, parallel, or even series-parallel, with the test coil. If a three phase winding is used, you also have a choice of wye or delta connections.
6. If you want to evaluate coil heating at this time, run the test setup for at least an hour under load, and check the coil heating.
7. When you find a wire size/turns combination that works, wind the remaining coils.

Marlec http://www.marlec.co.uk in England have been making micro sized (3 foot diameter) wind turbines using ferrite magnets and stators without overlap for twenty years and sellig them very well.  It was less than ten years back they went from single to 3 phase.  So coils that do not overlap are very practical and can be commercial.  Neos make for much smaller alternators but Marlec haven't noticed yet..

The overlapping coils with no iron I see on this forum very often make a very poor job of using the available space in a stator.  The idea of building a slotted plastic former disk that you can drop the coils into is ingenious but it takes up a lot of volume that I would use for copper.  That means the coils are slender and of course easy to overlap, but in the end I am convinced one would get better output (and it would certainly be less work) to fit simple big flat coils in a ratio 3 coils/four magnets or similar, like I do.

The maximum voltage is obtained by a coil-turn that has the same pitch as the magnet pitch (distance between centres).  Smaller coil turns have lower resistance but catch less flux.  If they get smaller than the magnet face I start to doubt if they are a good idea.  Their waveform will be poor because they only work when they are off a magnet face.

If a coil's maximum turn-size is the magnet pitch then the coils fit together to make a single-phase design.  (Coil pitch overall = magnet pitch) I dislike these because they vibrate, and also I like to have some turns that are larger than the optimum size, to occupy more space in the stator with copper (it's an opportunity for more power) and create a multiphase arrangement (less coils than magnets).  3 phase is the easiest to deal with in some ways but you could have 11 phases for 12 magnets for example and this would run very smoothly.

I have two approaches to connecting coils.  One is to connect all the starts together and each finish to a rectifier AC terminal.  This prevents parasitic currents and is very simple but rather intricate.  Lots of diodes and lots of wires.  The other way is to build a series/star 3-phase winding, by connecting every third coil in series to make a phase group.  This means fewer turns in each coil and thicker wire or more wires 'in hand' as you wind.

This stuff is all described in my windmill plans at http://www.scoraigwind.com/axialplans (now also available in French) and in my courses at http://www.scoraigwind.com/buildyo  The next course is in Costa Rica in February.