What I could find was that there are estimated to be about 350-500 million firearms in the US. The closest thing to a breakdown was 30-50 million semi-auto rifles, and probably close to a 60/40 split handgun to long gun over time (article shows trends). Murders per year using rifles and shotguns were less than 600 combined. Handguns accounted for over 6000.
So the number of hanguns is maybe something like 20% higher than long guns, but murders using handguns are 10x higher than with long guns.
Rifles are problematic as a single category since that involves hunting rifles. We can look at machine guns as a category that is neither rifle or handgun, and with limited use in hunting.
That said, the death rate of machine gun ownership may still be lower than handgun ownership, with a possible explanation being that it is easier to buy a handgun compared to a machine gun. A further jump into the data would then explore who are the buyer demographic of a handgun compared to a machine gun, and how machine guns compare to handguns in locations where both has similar restrictions on those who buy them.
An other avenue to explore would be to look at purchases of ammunition to determine which guns are actually in use for non-murder purposes. Hunting ammunition for example may show a much more common use than guns used for non-hunting purposes, and thus have a lower murder rate per shot. Machine guns and handguns might share a much more similar profile in that regard, possible normalized towards how much ammunition each use in a target range.
Relatively inexpensive plastic semiauto pistols, ruger is overrepresented, are purchased in small batches by people either pressured into it or who make their living marking them up and reselling to people who couldn't pass the background check. The majority of those sales go through a small number of known registered firearms dealers. The straw buyers are the ones breaking the law, but the dealers who knowingly selling to them aren't. There should be some policy fix, but the best I've seen is a state that limited purchases to 3 per month per person. It drastically slowed the flow into those groups driving up prices as old ones were recovered from crime scenes etc. Money and macro also has a good episode where they talk about how neighborhood beautification worked better than gun control for this type of violence. The scary rifles get used by the mentally ill because they look scary, and the large number of deaths at once combined with the inability to control your risk and the more privileged socioeconomic class of those harmed makes it more visible.
This is simply untrue, there are no full auto handguns legally sold to the public. Semiautomatic yes. And yes, some guns are way too easy to convert to full auto.
That said, the death rate of machine gun ownership may still be lower than handgun ownership, with a possible explanation being that it is easier to buy a handgun compared to a machine gun. A further jump into the data would then explore who are the buyer demographic of a handgun compared to a machine gun, and how machine guns compare to handguns in locations where both has similar restrictions on those who buy them.
An other avenue to explore would be to look at purchases of ammunition to determine which guns are actually in use for non-murder purposes. Hunting ammunition for example may show a much more common use than guns used for non-hunting purposes, and thus have a lower murder rate per shot. Machine guns and handguns might share a much more similar profile in that regard, possible normalized towards how much ammunition each use in a target range.