Sorry, but most of your ideas are far off base.
One is that this was a binary system, that a second star was behind the first at the time of the "pre-supernova" photo, and that they collided. Remember, they have very few photographs, are not using any data from space telescopes like SWIFT, and are therefore filling in the blanks.
Not correct -- they used both historical Hubble data to detect the star before it exploded and followup Hubble observations to confirm that the star has now disappeared. And they have data from the Keck Observatory with observations of the supernova. That's about as good as it gets for data.
We can assume that star evolution is moderately well-understood (though not completely), so if what they think is the input is inconsistent with what they know is the output, the chances are really good that the input is wrong, especially with such little data.
Star evolution is well understood for the bulk of the lifetimes of stars like the Sun, but there are still many questions about this sort of massive star. Such stars lose most of their mass during their lifetimes through stellar winds, which are themselves very complicated and not that well understood theoretically. And then the stars go through this luminous blue variable stage (which is what this star was before it blew up), and that is very poorly understood and is the subject of a lot of current work.
So it is in fact much more likely that this has uncovered a part of late stellar evolution of massive, luminous stars that is not correctly described by current models. We don't need any really bizarre explanation like iron planets falling into the middle of the stars. (And that wouldn't work anyway -- the planet would have to have a mass bigger than the Sun to have a big effect.)