Normally all you are after is twist restraint which isn't a 'force' it is providing sufficient torsional stiffness to prevent buckling. Sure the are some minor forces but in a normal system these will cancel out and resolve themselves as a group. In the case of blocking any lateral force will end up in the flooring.
The twist restraint is from member to member turns the local buckling problem of a member into a global buckling of multiple members and twisting of the twist restraints. All of which will take more energy (force) to instigate buckling.
Unless of course it doesn't....
EDIT:
I notice you are from Australia. I deal with AS4100. It is intently focussed on lateral restraint of the compression flange and doesn't at all address twist restraint of a member. The result is an easy recipe to follow that works for most structures but by ignoring twist restaint of the section in the absence of lateral restrain it kinda ignores the 'torsion' in lateral tortional buckling.