We use an accelerometer before and after a torquing operation(just to verify the torque wrench maintained it's calibration) in the nuclear industry. The wrench doesn't have to be the very best since the accelerometer sets up the wrench for success. We have to document equipment number, tool number, accelerometer number and values in work reports for traceability. I know most folks don't have access to an accelerometer but it suggests any adjustment of said wrench (without accel) may not ensure a dead accurate torque value. Perhaps over-kill for 'shade tree mechanic' but there is a possibility of over reliance on the wrench proper accuracy. Yes we are paranoid about everything lol. Like a previous post, consistency (without interrupting the adjustment) aught to remain stable.
