No. This shows a fundamental lack of understanding of statistics and empirical research. The research on gun ownership explicitly controls for known risk factors like location, income, crime rates, etc. That's what regressions are for, after all. After controlling for those factors that you mention gun ownership is, once again, associated with significantly higher rates of death by homicide and suicide.
I mean did you seriously think all of these highly trained researchers and statisticians were just running descriptive statistics?
Just for the sake of variety, and being absolutely fair (and also because the same point comes up in relation to completely different topics) I do think there are limits to what can be controlled for and there could possibly be unrecognised confounding variables.
Also, I think suicide is a different issue to homicide, and the two should be discussed separately. While it's true that sucide rates are clearly affected by the availability of means (e.g. if you have knowledge of and access to pharmacuticals you are more likely to kill yourself, same thing with guns) its still a different moral issue. I think those are two different arguments, really.
Of course there are also all the accidental shootings, particularly of other family members, plus lethal domestic violence. Seems like those things would obviously be far more likely with guns in a home.
Are those figures available, distinct from those involving people being shot by outsiders (a category which could, possibly, still be affected by unkown confounding factors)?