http://www.nhra.com/dragster/1999/issue02/Tech_Speaking.html
"Several misconceptions about valve springs influence racers to make poor decisions. A customer who says, "I don't need good springs because I'm running stock valves," is badly mistaken. Steel valves are heavy, and adequate spring pressure is absolutely essential to control their motion. A valve's inertia increases with the square of the engine speed, so even a small increase in rpm requires significantly more spring pressure to maintain valvetrain stability.
In fact, too little spring pressure is almost always the root cause of valvetrain failures. We spent a year studying valve springs using an Optron, a sophisticated electronic device that can precisely record valve motion and reveal valve float. We learned some shocking truths about valvetrain behavior at high rpm. Even with a relatively mild camshaft profile, the valves bounce on their seats before they close. If the spring is too light, the valve bounces uncontrollably. The valve hits the seat, rebounds, hangs in the chamber a while, then bounces erratically several more times. Imagine how hard this is on the valve and the rest of the valvetrain!
Another excuse I've heard for not using stiff valve springs is that they take more horsepower to compress. My reply is that each spring stores energy, and for every valve that is opening, another one is closing. Anyone who has been whacked by a torque wrench while turning a crankshaft can testify that the valve springs exert considerable force on the closing ramps! I have never installed stiffer valve springs on an engine and lost power; the improvement in valvetrain dynamics more than offsets whatever additional power is required to overcome the springs' resistance."