About that Australian "Gun Death" Study...

I can't help but note the graph shows a steep declining trend for better than 15 years before the '96 "buyback." I'm no statistician, but several questions come immediately to mind:

• If removing a fixed number of guns from the population resulted in a directly attributable decrease in deaths, can we also demonstrate a commensurate level in continued deaths for the guns remaining in the population?
• Which part of the population participated in surrendering weapons (i.e., "law-abiding"?) Are the death reductions exclusive to that segment?
• Has there been a commensurate decrease in criminal activity?
• Has immigration policy introduced a new and statistically significant population growth demographic for which gun ownership is nontraditional?

This is just off the top of my head--there are plenty of other questions that pop up, such as how they factor in higher suicide rates in totally disarmed cultures like Japan, why a temporary absence of a statistical rarity, mass shootings, is perceived as a cure against all future mass shootings, what the criteria is to define a mass shooting (more than 1? 2?...), and, to get theoretical (but it's justified based on longer-term historic "trends") what would the effects of a disarmed population be should the country either devolve into a Stalinist-style tyranny or be invaded?

I'd like to see someone who really knows his way around crunching the numbers and asking the right questions, like John Lott take a crack at this.