Yes it was me that made that comment.
It largely depends on the wind you have and the quantity of water you want to heat (or warm).
Dave has given you the information on the water side of things. 40degF is not exactly hot water but gives you a basis to work on.
You can easily calculate the energy in the wind. Do you know wind speed and how long it blows at a certain speed. What you can do in gales bears little resemblance to the average day. If you have wind 24 hrs a day you will do much better, here wind comes by 10 am and is gone by 7 pm except in storms.
You like so many others have totally misunderstood the issue of cogging. I have tried several times to lay this one to rest, but so many don't get it that they keep throwing it in and others believe it.
Cogging is a term that describes the magnet attraction of iron causing a torque variation as the machine is turned. Without iron there is no cogging, it is only a concern to motor conversion people and those using F & P motors. even then it is only a concern for starting.
Torque pulsating on load is not cogging and messing about skewing magnets and all the other tricks to reduce cog will not change things, except for reducing the output if you go to extremes.
All single phase machines have a pulsating load, it is inherent from the fact that the power in a single phase load has a sinusoidal form and twice line frequency.
If power delivered has a cyclic variation so must the torque.
If you intend to use a single heater then the load is single phase and by spreading it among the 2 phases of a 2 phase alternator all you are doing is to widen the coil distribution and to a point where you are not gaining much. Adding 2 voltages at 90 deg gives you little gain for twice the resistance.
This vibration from a single phase load is one reason not to do it but not the main one. Single phase alternators are common up to about 10 kW and if constructed robustly they work without trouble, but on a windmill the vibration will make you have to strengthen everything else as it is all a resonator.
The single phase winding only uses about 2/3 of the copper effectively and you need a 50% bigger machine for the same job. Losses are also higher.
To solve these problems you need a polyphase load. Loading your 2 phases with 2 separate heaters will be far better, but it adds complexity to the control. I would go the whole way and use a 3 phase load, it is not much worse for the control and has several advantages over 2 phase in use of connecting wire and the vibration may be marginally less.
Check your wind speed and distribution, decide how much water you need at what temperature and you can see where to go. If you have built the machine, try it with 2 heaters as load ( properly matched) and learn for the next one. If you haven't built it, don't.
Flux