I have to question any argument that uses 30-year-old data as its primary source; surely studies have been done since then?
http://focus.hms.harvard.edu/2002/Ma...y_control.html
This study, for example, comes from 2002, is also from Harvard, and shows a direct correlation between gun ownership and total violent deaths. Now granted, it's looking specifically at incidents involving children (accidents, suicides, and murders). If the original argument and stats were to carry over, you'd expect that as gun prevalence decreased and gun-related violence decreased, non-gun-related violent crime would increase. But this is not the case; non-gun violence does not increase, which refutes the idea that someone will turn to other implements when they do not have access to firearms.
edit: I just noticed this quote which actually speaks to the same sorts of conclusions that the first post makes:
"According to Miller, the danger of guns is not that they somehow cause people to act more violently--it is simply that people will be far more successful at killing themselves or someone else with a gun. "We see this in international studies," he pointed out. Forty percent of U.S. households own guns, and American children are far more likely to die from firearms than children in other industrialized countries. However, the overall violent crime rate in the U.S. is not much different from that in other countries. "What distinguishes the U.S. is its extremely high level of lethal violence," Miller said. "That's because guns turn assaults into deadly assaults, arguments into deadly arguments.""