50 turns would be a good starting point for 24v but with your wind you may want to reduce the turns a bit. Perhaps with N35 magnets you had better stick with 50 turns at least for a test coil.
You will need far heavier wire for the final thing, 2 in hand #15 would seem a good starting point if that is what you have. I think you could get thicker in.
"What should be the cut-in voltage" 24v DC Did you really mean that or did you mean cut in speed? With your wind I wouldn't aim for cut in below 140 rpm but others will tell you to go lower. With steady winds at 7m/s regularly I would raise it to 150 rpm. you know if your wind estimate is correct.
"What is the main factor for deciding to make a 12V or 24V or 48V machine. "
The voltage of your existing machine . If it is 12v I would think again, if it is 24 then that seems fine.
Tsr Well that is a compromise. High tsr results in fast running and a smaller alternator for a given output. It means noise, blade erosion and critical blade profiles for good results. Low tsr is quiet, blade profile is not very critical, starting is never an issue and the low speed means virtually no erosion. It means a large and costly alternator.
A nominal tsr between 6 & 7 seems a good compromise. You will not very often work at the design point, at cut in you will be up to at least tsr 8. On full load you will probably be below tsr5. If you drop below 4 with a nominal tsr 6 blade you will be likely to stall in the winds that you have frequently ( one reason I suggested a higher cut in speed).
The high tsr blades are best left to the commercial manufacturers with facilities to evaluate clever profiles and produce them in volume. For home build you don't want the trouble, uncertainty and noise. Very low tsr such as the Windflower work well but the low speed rules out direct connection with a sensible cost alternator.
Flux