Hey, Shannow, and others, here's a pic where I think you can really see how it works.
The inner, beveled, ring is the bypass. The inner ring of holes on the baseplate is very close to this inner bevel. Then, of course you can see the ADBV around the outside. The ADBV portion is considerably thinner, and also has considerably more surface area for the oil pressure to operate on. The baseplate also has much more area for the oil to flow toward the ADBV. The tendency, therefore, will be for the oil to push the ADBV portion out of the way and flow through the filter as long as this is the path of least resistance.
When the oil isn't able to flow as well as it should along the desired path, for instance if the filter material becomes clogged, the differential pressure increases, i.e., the pressure on the "clean" side begins to drop, the oil is then able to force its way past the bypass bevel which is able to flex inward toward the clean side.
I think it's an ingenious design, but I don't like the execution. I'm not sure what will happen as the nitrile rubber hardens. Also, like others, I think the spacing of the nylon cage is too wide. There's a picture referenced somewhere in this thread that shows a filter where a piece apparently blew out into the clean side. That's scary. I think that they're probably OK for relatively short OCI's, but I still don't think I'll use them anymore, especially when I can get a Motorcraft FL820s for under four bucks.