> come up with an artificial example where gun ownership doesn't result in increased violence.

Though I do not concede the use of the word "artificial" as correct in your sentence, bringing in the Swiss data point is incredibly relevant in response to a claim that "Actually, no, study after study shows that the amount of guns in a society is strongly correlated to its homicide rate, both directly as well as indirectly."