Try looking at it this way. If there is no alternator offset and the tail is directly in line it will point into the wind but you have no mechanism to cause it to furl.

Now it may be convenient to think of the original idea from which the furling scheme is devised.

Imagine the alternator with no offset but with a boom attached at right angles to the wind with a small tail vane mounted on it. Obviously the wind pressure on this vane will try to rotate the thing so that this vane goes down wind. The main tail will try to stay down wind and steer the prop as before, but there is now a balance of forces between the main tail and the little vane at right angles trying to get it out of the wind.

There is no force on the main tail vane when directly down wind so something needs to happen. As soon as the main tail is pushed sideways it starts to see a restoring force ( and this becomes a maximum when the vane is at right angles to the wind).

The result is that the thing balances with the prop at an angle to the wind where the force on the projected area of the main tail vane balances the force on the projected area of the control vane. As the main tail vane is significantly bigger than the control one it will settle with the prop at some angle to the wind ( say 30 deg). This balance holds for all wind speeds if you forget factors due to the prop ( too complex to consider now)

Early on someone realised that the side vane could be removed and the same effect could be obtained by simply offsetting the alternator axis. It looks better and gets rid of one bit of mechanism.

If you managed to follow that ( would have been better with a drawing), you will see that if the main tail is directly in line with the alternator axis the prop will end up running at an angle to the wind. It will not run in line as you assumed.

The only way to get the prop to run in line is to push the main tail axis in the other direction to a sufficient angle to compensate the alternator thrust so that the prop runs as intended. This is where the 10 deg comes from. In reality I think you will find that the angle required to do this completely is over 30 deg so it is a compromise between a bit of improvement and still looking reasonable.

Additionally requiring the tail to move more than 90 deg causes additional problems and when you get to 120 deg the thing becomes worse.

Fortunately as Roy said the drop in power with angle to the wind is fairly small up to about 30deg, so choosing 10 deg you end up running about 20 deg to the wind with little impact on the power.

What happens during the furling stage when the tail moves in relation to the alternator axis is very complex and I don't intend to continue to deal with that. You asked about the tail offset angle and I think I have explained enough to cover that.

With these schemes there is no way to balance the mechanical parts in a way that works before and during furling. The tail offset does tend to help this compromise but it is a secondary issue. Once it goes into furling the thing becomes totally unbalanced. Doesn't matter on land but makes this type of furling nearly useless for boats.