Yes, that is the page I am talking about.
You basically have 4 forces acting on the BOV.
-Pressure or vacuum on the top of the diaphram
-Pressure on the bottom of the diaphram via that lower port
-Pressure on the underside of the valve itself
-Spring pressure on the top of the diaphram
The combination of these forces is what determines the bahavior of the BOV. I'm trying to think of a short way to describe all of the typical situations, but I don't think there is one. The key point though is that the valve and diaphram are not the same size. The force on each is different at the same pressure since force = pressure times area. The diaphram is always larger than the valve, giving it more control over the BOVs actions than the valve. If the diaphram is twice the size of the valve, and there is 20 psi of boost (so 20 psi on the top of the diaphpram and 20 psi on the bottom of the valve), the valve will stay shut since there is double the force holding it shut. Also add in the spring pressure trying to hold it shut.
The problem is that the lower port allows the same boost pressure to reach the bottom the of the diaphram trying to open the valve. It effectively cancels the pressure on the top port, leaving only the spring pressure to hold the valve shut. If this was a static system it would still stay closed, but pressure in the IC piping (lower port) and in the manifold (upper port) is always varying, and at different rates. So the BOV tend to open at some pressure differential.
The purpose of the lower port, also called the fast release port, is to open the BOV more quickly when the throttle plate closes. Vacuum will build up on the top of the diaphram pulling the valve up, and pressure will spike in the UICP pushing on the valve (remember that is is ~half the size of the diaphram) AND on the lower part of the diaphram, opening it quickly. And in fact, putting atmosphere on that lower port will slow down the valves reaction time, and the valve/turbo become very noisy. And I agree that you would never want to cap off the bottom of the diaphram, as it would hinder movement of the diaphram.
This is where the solenoid switching comes into play. With the lower port at atmosphere you have both spriung pressure AND boost pressure on the top port pushing the valve closed, with only the smaller force on the valve to try to act against it and open the valve. It will hold extremely high amounts of pressure in this configuration. Most poeple stop testing at 35 psi or so. So at WOT (switch on throttle body) or under boost (hobbs pressure switch) the solenoid puts atmosphere on the bottom port, but when you let off it switches over and puts boost pressure from the UICP back on that port. DSMlinks nitrous controls can also be used to switch the solenoid. So you get the best of both worlds. Fast release when you let off, but excellent boost holding capacity at WOT.
Hopefully that helps to explain it.