The action of the wrist torque is a bit counter intuitive. It seems quite normal to assume that a positive wrist torque is required to obtain adequate clubhead speed. However this is not really the case. The wrist torque action can be a studied conveniently with a double pendulum mathematical model for the golf swing.
The wrist torque action is illustrated through three cases each with a different wrist torque. The first part of the down swing, duration of about .1 sec, is identical in each case (black portion of the Figures). During this .1 sec the club shaft is trying to back knife and a variable counter torque is introduced to model the equivalent of a mechanical dead stop, simulating the anatomical limits of the cocking action of the wrists.
From there on a small constant wrist torque is introduced, respectively a positive torque of +4 nt-m (Fig1), zero torque (Fig2), and a negative torque of -4 nt-m (Fig3). This forms the brown colored portion of the Figures. The "ball" position is indicated by the dashed line. The central torque is taken identical for the three cases.
Something quite interesting is showing up and is another example of a self regulating mechanism acting in a golf swing - notice from Figs 1, 2 and 3 that 'impact' is reached at the same time, 0.3 sec, Figs 4,5,6, notwithstanding there being employed a different magnitude for the wrist torque.

Figs 4, 5 and 6 show the 'clubhead' speed and it is evident that the action of the wrist torque is very minor with regard to 'clubhead' speed. Very likely a surprise for many but the negative wrist torque doesn't decrease but rather slightly increases clubhead speed. However the more important action of the wrist torque is to change the effective loft angle and approach angle of the club as can been seen from inspection of Figs 1, 2 and 3.
mandrin