The energy that goes into creating a wake with high surface waves is not available for propulsion.
My reasoning is as follows: I expect that at the surface a larger proportion of the work done (by the swimmer) goes to energy of the wake motion.
When the swimmer is sufficiently deep below the surface the water that surrounds the swimmer is closer to equal water pressure along the long axis of the swimmer. I expect that that will result in a more symmetric flow of water around the swimmer.
By contrast, at the surface the pressure distribution is the most uneven, allowing formation of surface waves.
See also:
Article titled: The science of underwater swimming: how staying submerged gives Olympians the winning edge Author: Anthony Blazevich Professor of Biomechanics, Edith Cowan University
Quote:
The breaststroke event was the cause of major controversy in the 1956 Melbourne Olympic Games as swimmers experimented with staying underwater for much of their races. The winner of the men’s 200-metre event, Masaru Furukawa of Japan, swam underwater for most of the first three laps of the four-lap race. This practice was swiftly outlawed after the games; swimmers were forced to surface before they could start to swim.
The breaststroke loses to the other strokes because the breaststroke swimming posture creates the largest waves. Hence the adapted breaststroke, (adapted towards optimization for submerged swimming), gave the submerged swimmers a significant advantage
Interestingly, Blazevich mentions that for the other three strokes remaining submerged only became a net win when swimmers became good at swimming in a dolphin-like manner. That manner of propulsion is described as 'Dolphin kick'. I infer that without a sufficiently well executed dolphinkick swimming submerged is not a net advantage (for backstroke, butterfly stroke, and fastest stroke).
