As far as I know or my best estimate is that the most invisible IRIS's are simply using derivatives and responses to plottings. This means you can't really accurately estimate the contrast effect or visible dynamic black effect by an On/Off reading of the IRIS by using a 0 IRE measurement.
A good test might be measuring both a JVC and Sony with a pixel fade-in pattern that starts at a full black 0 IRE then adds 1 pixel at rgb(1,1,1), then 2 at (2,2,2) on opposite sides of the screen, etc... Slowly going up all the way to multiple white pixels in the image at various parts on the screen. Take the measurements of the black level at the farthest point from the pixels in succession (middle of screen) and then back to full black and then to full white and see the effect of the on/off as it changes (well wouldn't technically be on/off, since you'd be switching between full white and partial black with pixels, but you get my point), etc... Something like full black, black with 1 pixel at 1,1,1... Full black measured to black with 2 pixels at 2,2,2, and keep going, could even throw in some full whites back to those patterns, or all 3 in a row in succession. Average it all out or something.
I'm sure someone could come up with a better test, but I mean something in this general range.
I am getting somewhere around 50,000:1 Native On/Off from a JVC RS-45 at farthest throw, if I raise the manual aperture from -15 to say -5, the loss in Native on/off is not terrribly visible (but you can see it). So I'd say even if the IRIS has a small effect starting at 20,000:1, you might not notice much black level loss compared to a JVC. Maybe vs. an RS-55 or RS-65 a bit, but the thing is how many people are actually getting 100% of the Native from their JVC's (I am but I am using an HP screen at farthest throw), so most of the people buying this projector would probably be doing so for the extra lumens and 4k ability with a larger screen. The black levels should be good enough in the overwhelming majority of scenes.