The picture above of the graph mentions it is at peak TQ, should be obvious that as scavenging goes down at lower RPM then there is less negative pressure at the lower lifts. Even above the TQ peak it is probably less. On an engine with tuned intake and exhaust runners I bet it changes with the helmholtz resonation. IE the magnum beer barrel intake is tuned around 1800 RPM to build pressure behind the intake valve as it opens combined with the negative pressure building in the cylinder makes for a big pressure differential then add on tuned headers for that RPM and you could get very very low pressure, but on the other hand there won't be much scavenging at that RPM if the exhaust is tuned for 4000 RPM. Also if the cam is wrong the scavenging RPM shifts around changing when the pressure differential is greatest. Even on a more finite scale the resonance the pulse may not hit the back of the valve till it is at max lift or even as it's closing. I would say RPM is definitely going to change the depression level and timing.

Probably need a super computer to get it near perfect for wide RPM ranges and throttle opening angles. If you were tuning for a single RPM it would be a little easier