OK . . . . stick your arms out from your shoulders parallel to the ground . . . .
Now make your hands move parallel to the ground in a cicle by turning like you were a helicopter. . . . your spine is ONLY required to twist to keep them on that horizontal plane . . .
Now do the same thing only add waist bend to get your helicopter dealie moving on an inclined plane . . . your objective to keep your hands and arms moving in the same inclined plane . . . to do this notice what your hips and spine have to do . . . . there is a lot of stuff going on below your shoulders to keep your hands moving in that orbit without disrupting it . . . and alot of it is hips sliding and spine flexing and extending . . . . on the way back and through.
When I mentally envisage an axis of rotation in the downswing, then I actually do imagine that the axis of rotation is a diagonal line drawn between the left foot and the upper swing center. In other words, I can more easily imagine an axis of rotation that is diagonally aligned rather than vertically aligned (and which is also centralised between the feet).
On page 53 of his book, VJ drew a diagonally slanted line going through the torso - that follows the diagnonal line that I just described.
However, then on page 37 of VJ's book - figure 11 shows a pivot axis line (in white) at impact that is vertically oriented, going from the left inner foot to the left shoulder socket. That's why I don't know what to make of VJ's pivot axis concept, where the pivot axis apparently changes its axis of rotation during the downswing.