I'm sorry, but this old racer's tale just isn't true. Golf balls have dimples in order to "hold" the transition zone from laminar to turbulent flow further back on the sphere, by inducing turbulence around the ball. This reduces the ball's flight drag by shortening the separation region of vortices behind the ball, which create most of the drag of the ball's flight. Your ports are designed for laminar flow all the way through -- there is no separation zone here. Obviously in a real port there will be some turbulence, but separation zones of the sort we're talking about require a "dead spot" like the area right behind the valve stem. Perhaps roughing up the valve stem will help, but not the port surface itself.
Speaking simply, it makes no difference.
The "velocity profile" you're talking about IS the boundary layer, by definition. And if dimpling laminar surfaces would reduce drag, we'd see some really dimply-looking 747 fuselages. Dimply-looking space shuttles. Dimply gliders. F1 cars. The list goes on.
I've also heard that rough intake walls encourage fuel droplets to atomize. I don't know anything about the physics of atomization, or whether that's a problem that needs to be worried about or not.
If there's a real aerodynamicist in here, feel free to correct me. I used to design the electronics for unmanned aircraft, and when doing my Camaro's head I asked our resident aerodynamicist his opinion on the polishing "wisdom", and that's what he told me.