As 4 pole is more common this is what all spinning in my head. I have drawn it to scale.

36 slots rotor
If I go for 8-10-12 winding, 4 pole winding then a pole width will be equal to 7 teeths in stator as in Zubbly's work.

Teeth to teeth  = 10 degrees (36 slot).
7 teeths = 70 degrees
Arc lenth at 230 mm (73 mm dia. rotor after turned down)
= 70/360 X 230 mm (circumference)
= 44.75 mm

According to this image I will be able to put
8 x 12 = 96 magnets over the rotor.

The total magnetic volume will be equal to no. of magnets x Pi x r^2 x h
= 96 x 3.14 x 6.5 x 6.5 x 10
= 127358 cu. mm. of magnets OR
= 7.771862 cu. inch
So according to Zubbly 150W per cu inch rule, I will get
7.771862 x 150 = 1165W of power.
This is more than Danb's 10 footer which i thinks in 1000W.

To Flux,

As Sparweb has said and what I thinks

"Changing the number of poles will not affect this result (much).  Think about what frequency of AC you want to receive - if it doesn't matter, then go for whatever you think it easiest."

Increasing the No. of poles will only result in increase in frequency.

Secondly if 6 or more poles will result in more than 10% in power output, and if I get some pointers (links) on coil winding and connection I will go for it.

One more thing that baffles me is that there is lots of dead copper?(copper that is not in stator slot) in 8-10-12 winding. Am i right on this? Is there an more good alternative to this 8-10-12 winding? I also have a thought that there must be a more good alternative that this 8-10-12 winding.

I like to have more pointers on winding.