"First I found myself a suitable surface such as linoleum . Next I took off my shoes and socks and stood next to a wall for support . Then , taking most of my weight initially on the right foot which was kept flat on the ground ,I put my left foot into toe off position so that only the ball of the foot and the toes were in contact with the ground . Then I carefully transferred some weight onto the left foot and finally,with the ball of the foot and toes under some pressure , rotated the foot outwards producing torque . I found that the toes played an important part in resisting the outward rotation and that they also began to become spaced out from each other possibly engaging the adductors .( please note that I am not suggesting that anyone copy the above exercise sequence or injury may result )
So perhaps the toes can significantly aid grip on a flat rigid surface ? "

A final point related to the above quote , taken from my last post , might be of interest .

Several studies have looked at the control of angular momentum through the coordination of body segment movement during walking . I feel that it is likely that in the unshod condition ,where walking or running produces excessive torque , the subsequent outward rotation of the foot will necessitate increased eccentric activity in the toe adductors , such as the transverse head of the adductor hallucis , to control toe separation . It is also plausible that stretch receptors in these muscles will send proprioceptive afferents which might then help produce adjustments to body segment movements ,such as arm swing , to counterbalance the torque being produced in the leg and foot and so produce more efficient movement with less internal , torque related , joint strain .