Torqueing knock sensor/crow's foot

[Flanner, Chris]

Here's how I see it . . .
 
Changes in the location of the crow's foot relative to the socket
connection of the torque wrench (I'll call it the "head") has no effect
on the reading as long as the head and the crow's foot apply torque in a
parallel axis.  No matter where the torque is applied by the crow's foot
(or socket or whatever) the equal and opposite torque is seen at any
desired point in any parallel axis including the torque wrench's head
where it is converted to an indication.  Only the force on the torque
wrench handle and the equal and opposite reaction force at the fastener
change when the position of the crow's foot changes.  This, of course,
ignores the effect of friction on torque readings but that's another
issue.
 
Am I wrong?