Read a scientific paper once that suggested that the Meteor Crater in Arizona was actually one in a string of about 7 craters in different spots all over the world. They used various half-lifes to date them, and iridium content as well. This crater dates from one of the earlier extinctions, not the relatively puny Dinosaur killer. The interesting part about the paper was that they used a computer model to simulate the backtrack of the tectonic plate drift. There were five in an exactly straight line, and three more that were extremely close to a straight line. The inference was that it was a broken up comet, similar in form to Comet Shoemaker-Levy 9, which strung out before colliding w/ Jupiter.
The Death Star idea is a very good one, and many of the new huge telescopes are being designed with the idea of looking for such a nearby, dark, cold star. Very tough to do, however. Theoretically, such a star would still produce neutrinos, as well, and our neutrino detectors are showing a great deal more than we were theoretically supposed to find from the sun and the known nearby stellar sources. Just a thought....
One last bit, the explanation I heard about the galactic dust explanation wasn't that we wobbled up and down through the disk, but rather that the sun orbited the galaxy faster than the arms of the glaxay did. Between the arms lay vast stretches of interstellar gas and dust, which would necessarily constrict the heliosphere (ie, the boundary between the edge of the solar wind and interstellar space) to inside the Oort cloud. This means the Oort cloud would be given extra energy in the form of collisions with dust and gas particles, which can result in many of them colliding and being ejected from, or falling towards, the solar system. The infallers being the big problem, of course. And, again, our progression through the bands of the galaxies seem to work on the same time scale as our major extinction events.
Finally, the extinctions don't always happen every time we hit one of these bands, but they do almost always occur with the same period; this is why there's a discrepancy between the number of extinctions and the amount of total time between them.
Or, at least, as I understand it.