It is easier to track distributing a sold product. It is very difficult to track what a person makes in the comfort of their home. I dont think the federal government keeps a registry for the purchase of milling machines.
If everyone who wants to own (to keep using this one example) an AR-15 had to buy a CNC, that would be quite the commitment. If the current model of people buying many milling machines and selling their services is the only viable way to avoid that, then you've already seen the feds start to take notice. After all, if you can get word of mouth around enough to make money, you're making enough noise to be heard.
3D printing is presumably going to become ubiquitous, and require no skill or experience to operate and maintain.
But that's the thing - with the present level of technology, you can have a device that is, essentially, pre-programmed CNC to make AR lowers (which is what Ghost Gunner is, for example) for under $300.
To make a fully functional firearm from there, you also need a complete upper - can be had for another $400 or so these days; and a complete lower part kit with buffer tube and stock - can be had for less than $100. For a total of $800, some assembly required (or you can pay another $50 and have pretty much any gunsmith put it together for you).
So basically, a person can take GG and produce an unregistered, untraceable, no-background-check-required AR with it for $800-900. For comparison, the Bushmaster AR that was used in Sandy Hook shooting costs ~$700.
For a cartel? It's nothing. For most people who want a particular gun, and don't intend to monetize the purchase, never mind the time taken to learn how to work and maintain it? That's a hell of a thing to be honest.
The real barrier is the knowledge, you don't need a relatively new CNC machine like that if you're just going to make a few for yourself. And old but high quality non-CNC machines are sufficient for that, and require a somewhat different set of skills.
But that's not the scenario I'm thinking about, nor is it relevant as of now in a country which has more than doubled its gun production since Obama was elected, with > 100 million being sold in that period (obviously not all those were new).
If the barrier to a decent firearm is knowledge of how to work and maintain and older CNC, that seems like a pretty excellent means of limiting firearms. "You're free to have what you can make from scratch, up to a certain point."
The cost argument is sound, and I'm not disputing it, but rather I'm thinking in the broader terms of the "Gun Control" debate. At some point in the future, maybe a decade, maybe two, supply-side gun control is going to be a pointless exercise. My argument however, is that the time has not yet come, although it's clearly on the horizon.
For now "Just get a CNC mill and learn how to use it" is still a substantial hurdle. I think it's fair to say that it's a hurdle that Adam Lanza wouldn't have managed to overcome.
Gun buybacks are the best gun control method I can think of. supply side is problematic especially in the US, where the amount of owned, unregistered guns is enormous. I think the first step is reducing the supply of used guns. and, it has the side effect that the people who sell guns to government for cheap, likely would sell the gun for cheap to anyone. Might as well be the government, though.
In the US at least, there are three classes of people who are willing to sell guns cheap at a buy back, and this is why the concept has never taken off, along of course with the cost, which isn't effective at buying votes:
Criminals wishing to dispose of weapons used in crime.
Savvy types who sell guns that are worth less.
The ignorant/those who don't want to be bothered to easily dispose of guns that they've inherited or whatever.
The savvy types will work the line to buy guns from the latter that are worth significantly more than the price.
3D printing is presumably going to become ubiquitous, and require no skill or experience to operate and maintain.