Because their Reynolds number is very big and their boundary layers are already turbulent.
The story is so oversimplified that raising questions from it is just pointless.
The facts are as follow:
1. Roughness tends to increase drag because makes boundary layers turbulent.
2. Turbulent boundary layers do stand higher adverse pressure gradients prior to separation
3. Separation increases drag much more than turbulent boundary layers.
Then, there are some applications where you would have a separated flow, and promoting turbulence through roughness would reduce the drag. This is not the case of aviation. It is not the case for sure of sharks when they are not moving their tails. It may be the case of sharks when they are moving their tails to obtain propulsion.

well, not necessary, since submarines Reynolds number (rho*V*L/mu) tends to be much bigger, and it may happen that the boundary layer is already turbulent.
I'm confused with the story itself, it is poorly written.
From my knowledge on fluid mechanics, turbulent boundary layers increase drag unless separation is avoided (turbulent boundary layers stand higher adverse pressure gradients prior to separation). I guess that what happens with sharks is that they can twist their tails with higher energy providing higher propulsion because the boundary layer withstands such pressure gradients generated thanks to the triggering to turbulent (which should not be the case because their low Reynolds number).
Of course, higher roughness would induce higher drag, so the evolution has found a right equilibrium for them.

Well, perhaps it depends on your distro. Some shells could have it built-in.

rsync -e ssh -auvzH --progress MyDir/ someuser@somehost:destiny/dir/

real UNIX admins don't sudo

Specially Spanish real UNIX admins, since sudo means "I sweat", and real UNIX admins do not sweat just as Chuck Norris doesn't.