You are correct, the only force that could be pushing the LCA backward on the pin is through the strut rod. The strut rod is literally the only thing that locates the LCA forward/aft, that is its only purpose and no LCA bushing regardless of kind does that.
The order in which you torqued the strut rod and LCA pivot nut shouldn't really matter. Even if you torqued the LCA pivot pins before the strut rods it would just be the compression distance of the strut rod bushings that would push the LCA backward. And that shouldn't be enough to separate the LCA bushing because during driving operations the rubber strut rod bushings extend and collapse, the LCA bushing flexes with them. So if the full backward extension of the strut rod and bushing caused the LCA bushing to separate, it would have done that after driving the car too.
To me, as long as the strut rod itself isn't too long, then this has to be a manufacturing issue for the LCA bushing. The only thing that holds the rubber LCA bushings together is the force from the compression of the rubber bushing between the inner and outer shells, so there really isn't much to think about. Either that compression force wasn't high enough because of poor tolerances, or the strut rod is too long (it would have to be for a different car, or the wrong year).
