It should be noted that this chart compares gun homicides to gun ownership, which... of course those will correlate
If we plotted kangaroo injuries vs kangaroos per capita, we'd see a similar outlier in Australia
It would be more useful to see gun ownership compared to total homicides, to see if an overabundance of guns correlates with more murders. Even then, though, a correlation between the two might not be casual in that direction. It may instead be that in areas with a high homicide rate, people are more likely to own a firearm for defense.
What you would need to prove is that places with high gun ownership have significantly higher homicide rates, but places with high homicide rates don't have significantly higher rates of gun ownership