The extremely obvious, plain reading of the gun ownership / gun murder rate data is that there's no relationship. Check this scatterplot[0].

The studies in your link are after "controlling for poverty and urbanization", "after accounting for rates of aggravated assault, robbery, unemployment, urbanization, alcohol consumption, and resource deprivation", etc.

I'm not a statistician enough to argue the methodology of the papers, but I'll say given that the trend only emerges after lots of adjustments, it makes me a little skeptical. At least, it would be pretty easy to let some bias or motivated reasoning slip in.

[0] https://en.wikipedia.org/wiki/Gun_violence_in_the_United_Sta...