My take on this is when the valve is closed, pressure on the back side of the intake valve opposes the valve spring pressure. So, when the intake valve closes it can bounce off the seat due to the supercharger pressure detracting from the valve spring pressure that is trying to keep the valve closed. This effect will show up on the dyno graph just like a valve float condition. On the dyno with my supercharged FE running 17 psi of boost, I lost about 400 usable RPM going from naturally aspirated to supercharged, and I think it was due to that effect, although there's no way to prove that. But when I increased the intake valve spring pressure by 60 pounds, I got the RPM back.
The way I calculated how much additional pressure I needed was to figure out the area of the back side of the intake valve and multiply that by the peak boost pressure. I have 2.25" intake valves in my supercharged engine, so neglecting the area of the valve stem the area of the back of the valve is 1.125*1.125*3.1416, or 3.97 square inches. So, using about 4 square inches and 17 pounds of boost I figured I needed to add 68 pounds of seat pressure to the valve. I added 60, and that solved the problem.
I'm currently revamping that supercharged engine and will be running it on the dyno again sometime in the next few months. I'll see if I can repeat that experiment when I do so, and check to see that the results are the same.