The first thing I want to know is if they are distinguishing between legally-owned and illegal guns. In a 'region' like New York City, for example, most of the guns are illegal, and in the hands of the gangbangers and criminals who comprose most of the gunshot victims. I'd expect a strong correlation between 'gun ownership' and victimization if they counted an illegal weapon as an "owned' gun.
If you only count legally owned guns, and if you use an unbiased way of choosing your study areas, you'll have lots of data from places like Vermont, with very high gun ownership and very low crime. Most of the high-crime areas whould show very low (legal) gun ownership, leading to a very, very different conclusion.
I'll reserve judgement until I see the facts, but at first glance this looks bogus.