That is the intriguing conclusion of a new study:
Michael H. Morgan and David R. Carrier, Protective buttressing of the human fist and the evolution of hominin hands. The Journal of Experimental Biology, 2012. Freely accessible → LINK [doi: 10.1242/jeb.075713 ]
The derived proportions of the human hand may provide supportive buttressing that protects the hand from injury when striking with a fist. Flexion of digits 2–5 results in buttressing of the pads of the distal phalanges against the central palm and the palmar pads of the proximal phalanges. Additionally, adduction of the thenar eminence to abut the dorsal surface of the distal phalanges of digits 2 and 3 locks these digits into a solid configuration that may allow a transfer of energy through the thenar eminence to the wrist. To test the hypothesis of a performance advantage, we measured: (1) the forces and rate of change of acceleration (jerk) from maximum effort strikes of subjects striking with a fist and an open hand; (2) the static stiffness of the second metacarpo-phalangeal (MCP) joint in buttressed and unbuttressed fist postures; and (3) static force transfer from digits 2 and 3 to digit 1 also in buttressed and unbuttressed fist postures. We found that peak forces, force impulses and peak jerk did not differ between the closed fist and open palm strikes. However, the structure of the human fist provides buttressing that increases the stiffness of the second MCP joint by fourfold and, as a result of force transfer through the thenar eminence, more than doubles the ability of the proximal phalanges to transmit ‘punching’ force. Thus, the proportions of the human hand provide a performance advantage when striking with a fist. We propose that the derived proportions of hominin hands reflect, in part, sexual selection to improve fighting performance.
I wouldn't dare to comment much but for what I have read in the paper, it looks plausible, notably because it is indeed significantly efficient versus the open hand (as much as 3x) and because chimpanzees can't do it... but australopithecines could.
However I can also imagine this development as a side-effect of other adaptive uses of the hand, such as grabbing a spear, which is no doubt a much more daunting weapon than a naked fist... in most cases at least.