I think Mr Bruce is talking more about the FFR adapter bracket itself, right? Solely based on that, he's right: pass a rod through the upper ball joint hole, and it passes forward of the lower ball joint hole. The spinde also angles the bracket forward, as viewed from above (see GatorAC's picture).
However, that's moot once the car is properly aligned: a line drawn through the upper and lower ball joints to the ground falls forward of a line drawn vertically to the ground through the spindle - that's positive caster. The relative angles of the ball joint studs are irrelevent (up to the limit of their travel) because they can pivot in any direction.
IMO, the reason for the angle on the upper ball joint is the original Mustang geometry: strut. The upper pivot in that case is the top of the strut tower. Let's assume that both an FFR and a Mustang use 3 degrees positive caster. Viewed from the side, the FFR upper pivot (ball joint) at ~10" above the lower ball joint is indeed farther forward than the Mustang upper pivot (strut tower) at ~24" above the lower ball joint - but the caster angle is identical, and in both cases the line to the ground falls forward of the axle centerline to the ground. So, Ford had to move the spindle-to-strut attachment point as far forward as possible to keep that angle to the top of the strut.
Clear as mud, right?
