It's quite simple really. The dimples improve the aerodynamics of a golf ball and reduce turbulence in the balls wake; reducing drag. Nodules are not aerodynamic and would increase the turbulence. 

http://golfnews.homestead.com/golf_ball_dimple_dynamics.html

I'm not sure "going further" is the design aim.

A plain ball gets a laminar flow over the front half which detaches chaotically on the rear side. The vortices make an oscillating pull on the ball which gives it a slightly random flight path. The Beckham Bend makes the ball spin so the surface layer pulls these vortices to one side of the ball, making them all act in the same direction and dragging the ball in a curve.

The puckers on a golf ball (whether protruding or inward) are there to break up the surface layer flow and make micro-vortices which average out on a much smaller scale, making the ball fly much straighter and more predictably. (Goal width 8 yards, golf hole 4.25 inches.)

I think protruding surface elements would be more likely to make the initial club contact unpredictable (especially with ceramic clubs) whereas dimples provide initial contact across a larger surface.

Going further is also the design aim though... for a smooth ball, as the flow detaches from the surface just after the mid point, the vacuous area left behind creates a negative pressure gradient compared to the high pressure at the front, resulting in a force acting against the direction of travel.

The fact that you can get different ball types depending on spin, trajectory or stability suggests that different dimple shapes will give you different results. The point of the dimples is to create a bit of turbulence in a controlled way allowing the boundary layer to stick to the ball further round than on a smooth-surfaced ball. This results in a smaller pressure gradient between the front of the ball and the back, minimising the force acting against the direction of travel.

It can't be that the dimples minimise turbulence as the ball is constantly spinning, and most likely with back spin, which would cause a very quick and steep rise. Any reducition in turbulence in this way would also make the ball act like a swinging cricket ball. Any small amount of side spin off the swing would cause the ball to pull or hook drastically, which isn't the case.