Everything with wind is a compromise and everyone's compromise is different so there is no correct answer.
The requirements for someone with skill and engineering facilities and a bit of money will be very different from someone in a third world country.
Unless you are designing for mass production in a competitive market the thermal modeling stuff is interesting but dreaming.
With mppt the alternator losses even from the conventional alternator used here fall to quite a low level, it is easy to cope with a loss of about 100W, but dissipating 500W plus in a small alternator is a challenge.
Your concept is the easy one, keep efficiency high and not worry about clever cooling and thermal modelling. There really is no other way to get high overall electrical efficiency with battery charging than with mppt.
If you have a limited budget, limited resources and no means of repairing complicated things then your only requirement is that the stator does not burn out, the loss of overall efficiency is better solved with a bigger prop, it won't need a big increased in swept area to offset the efficiency loss.
With the simple scheme some chase every form of cooling and it does help a bit, but in reality you can't dissipate kilowatts in any small stator. Far better to use a bit more magnet and copper and add the necessary loss in the cables rather than include the necessary loss in the stator. Finally if the furling works within the alternator limits it won't burn out. If the furling doesn't work it will fry sooner or later.
Now back to your proposals, it will work great for you, you know the requirements and have the facilities to do it.
Even an alternator designed for direct charging with stall operation will become much more efficient with mppt. To get the absolute maximum out of mppt you can push the alternator efficiency way up on what is possible with direct charging, although the heating consideration is no longer an issue.
Now the speed increasing transmission is another trade off. Once you go to ferrites efficient low speed alternators become big and heavy and a bit costly, if you can get them up in the air there is no other real problem.
Increasing alternator speed is the only way to reduce size and weight for a given efficiency. With iron cored alternators there is a big snag with start up and low wind performance and it may not pay off for small machines. With air gap machines there is no iron loss and bearing losses and chain loss will not cause any starting problems and unless the speed increase is stupidly large you won't see significant low wind loss. You end up with a large gain in alternator efficiency for a given size and the overall set up should be lighter than the direct drive equivalent.
Chain drives do work when properly engineered and are a good option, belts and gears really are best avoided for wind power.
If you follow what Chris Olsen has done you will end up with something reliable and very efficient. Ferrite magnets have a real advantage if you are in an environment where neo corrodes,
The only thing I question is the point at which you go to chain drive with a 12v system, for such a low voltage system I personally would go for direct drive even with ferrite. It is difficult to hold down big machines at 12v, a 12ft mppt machine will require lots of batteries, although you do at least get away from huge rectifiers and connecting cables. For 24 or 48v the chain drive starts to look more attractive.
Why all this rambling? It was provoked by the thermal article, not by your proposals. I just wanted to set the field straight by explaining what suits one constructor would be useless for someone with different needs and facilities. Understanding cooling is good but the requirements are wildly different in trying to push an alternator from 95% to 98% compared with trying to cool something running at 40% efficiency. Once you choose the mppt route you don't need to worry much about cooling, you can still get more out by chasing efficiency but cooling would only be an issue when trying to use an alternator way too small for the job.
Flux