The lower mount does indeed have an indentation for the spring to sit in. The spring end fits perfectly in this groove, and I made sure the end of the coil was butted up against the end of the groove as well.

When I raise the car and extend the strut assembly, I'm actually able to reach in and twist the spring with a forceful jerk, rotating the bottom in the lower mount myself. I can then jerk it back into position. The top of the spring always stays in place in the rubber upper mount, although that's probably because that mount is designed to rotate in the bearings of the backplate.

I have the star nut tool from IPD, and tightened both it and the large nut on top after assembly to what I approximated to be 50 lbs. However, I was under the impression that would just eliminate the vertical play in the assembly. How could tightening that upper bolt decrease how much the spring wants to rotate, especially if it always freely rotates in the top assembly?