A) The study is worthless, to be perfectly frank, in terms of it's control group. There are how many gun-related crimes in America each year, and the control group is 65 over a 40 YEAR period? Anyone who deals in statistics on a regular basis knows that is a really bad control group.
B) If they have included Columbine, Red Lion, or any other school shootings (and I'm making an assumption here), then the description of "legally owned" is utterly incorrect. As I said earlier, it is, as far as I am aware, a crime in all 50 states for a minor to possess a firearm. If they include a kid taking daddy's gun, that is NOT a legally-owned firearm, it is a stolen firearm.
C) Compare the majority of murders in the US committed with illegal firearms to this really poor study. You'll find that most homicides committed in the US with firearms are committed with illegal firearms. Yeah, the corner crackhead may not go on a shooting spree and take out thirty kids (he'll just hit one or two), but the fact of the matter is that most crimes committed with firearms here in the US involve illegally obtained firearms.
D) It's chosen to analyze a specific subset of gun-related crimes which, to be frank, constitutes a small percentage of gun-related crimes. If it's just examining mass shootings, of course the stats look bad. But when you consider how few of these there are, it's not such a big number.
People are running around talking about crazy killings, forgetting that these perps make up a very small percentage of gun owners. It's the illegal guns which are infintely more dangerous.