They appear to have two different air bleed rates controlling the vacuum. That could be a function of the bleed hole internal to the valve. I do not know for sure, but this may be a function of whether it is a positive backpressure valve or a negative backpressure valve. That could suggest that manifold vacuum may vary with the transmission type. Just speculating - I honestly do not know.

The factory service manual (FSM) has a good cross-section view of the EGR valve and some info on how it operates. I have downloaded the 1996 FSM, which should be same as 1997. See page 6E-939.

Don't know if I have given you this link before:

http://www.mediafire.com/?40mfgeoe4ctti

 

Interessting scheme
So the valve isn't meant to stay open when vacuum is applied and the port is closed by hand like in our test. Air will enter through this bleed hole, right!?

Funny thing is that we tried the same at a Lt1-Caprice and there the vacuum remained... hmm

 

found a document with a good write-up about the function of the EGR-valve itself.
I assume that the valve should stay open as long as you apply vacuum on a non-running engine.
Funny thing is that with a fully warmed up engine the replaced valve seems to work, means: I drive this same street. One time the temperature reaches operating temp (~70°C at the gauge) just at the end of the stretch with slight incline and BAM! there comes the light. But if I drive about 10 miles, operating temp at ~80°C and then up this road, no problem...