Not sure there is much to add to this. New ideas are always welcome and in many cases new ideas are nothing more than re-invention of old ones.
Much depends on your wind resource and how much you rely on very light winds. There is no doubt that you can do dramatically better in high wind areas if you depart from things designed to work best in winds near cut in.
There seem to be several issues here as I have said before, there can be no doubt that the air gap axial machine designed for cut in at 7mph will be restricted in maximum current capability in high winds because of the inherent lack of cooling. You can devise methods to improve this and similarly it is probably easier to achieve better cooling from radial designs. If you need to make full use of winds above 30 mph and can use these high currents then modifications are likely to be a good idea. Efficiency and current capability are not the same thing and if you use one machine to load from cut in in light wind right up to producing high power in high winds then efficiency has to be lost somewhere. If the efficiency loss has to be in the stator then some better form of cooling is necessary but the inefficiency will still be there. Any attempt to raise this efficiency will stall the blades.
There is probably a strong case for a machine that will survive much higher outputs if it fails to furl, that seems to be the main virtue of motor conversions but no way are they more efficient, they get rid of the internal heat better but you get less effective power into the battery but it is again clouded by the different type of blade matching.
Efficiency is again tied up with duty cycle and you can manage a far higher duty cycle with a high efficiency alternator but the lower your efficiency the better you will need to cool it.
Wind doesn't give 100% duty cycle or anywhere near it so as long as you stay within the limits of the alternator it doesn't have to be rated that way. Commercial large alternators for high speed engine ratings can't be run at the low efficiencies of these wind machines for the duty cycle they use. You can go so far with forced cooling even using hydrogen but in the end the losses have to be kept small. Their type of loading is impossible over a speed range into a voltage clamp.
In the end I suspect it is suitability for the application rather than efficiency that we are really chasing and if you want lots of power from a small alternator then there are indeed many ways worth following.
If you can do better with a radial or with a delta winding then that is fine as long as it doesn't impact on low wind performance if you must have those few watts in light winds. If a few watts on low wind days are not important then you can drastically improve the existing designs by going for a higher cut in speed. In fact in the end speed is the fundamental issue here, whatever you do the machine becomes bigger heavier and more expensive for a given rating as you drop the input speed. Before the days of neo magnets low speed alternators were massive for a given rating, so much so that gearboxes were common.
As Hugh and Dan have implied the slotted iron core is not the best approach for best low wind results. I have no experience of powdered cores, they made no inroad in conventional machines but may offer some promise here, the loss will certainly be lower, it remains to see what flux densities they will carry and how much magnet you save over an air gap design and whether the old reactance limiting comes back to haunt you. I suspect it may be a viable idea as long as you avoid teeth. Once you avoid teeth then iron cored machines are very satisfactory with core losses way below the conventional.
I look forward to Chris's experiments, I know it has been tried before but I have seen no really useful figures to base anything on. Radial will be easier to wind effectively and easier to cool. As for the choice of star or delta that is a personal one. For machines operating on mains there is no real difference in performance, motors are available for choices of voltage with star or delta with identical ratings.
There are other issues when feeding a clamped rectifier load and conventional practice may not have been evaluated by many people. It does seem strange that the biggest market for battery charging alternators has virtually standardised on star, I would have thought the car alternator market would have brought to light any real virtue of delta if there was one, but again a car alternator is built for cost size and rating and efficiency is very secondary, a bigger fan is cheap .
There has been a lot of interesting discussion here, no doubt useless to the person who asked the question and as usual there is no real conclusion except that high wind machines may not be the same as low wind ones. I can't contribute any more so it stops here as far as I am concerned.
Flux