Regarding the first paragraph, the way they measure this is observing the incidence in different circumstances. Similar country but higher punishment? See if fewer people do the crime. Oversimplified.

The research shows that the deterrence effect exists but, beyond a certain punishment level, it doesn't do much anymore. What helps is primarily the odds of being caught at all (and then an appropriate punishment) and secondarily the time between violation and punishment (I didn't realise this would matter for adults but apparently so)

Going to jail for two years, four years, or six years, either way you lose your social life, the roof over your head (once you get out), your job, everything. It's a doubling or tripling of the sentence but is it really that different? If I'm okay incurring 2y prison sentence... I'm probably not the target audience for this but I imagine such a person would also risk 6y if they want someone gone that badly and the odds of being caught are low enough

Prevention is golden, as you say. But then rehabilitation is silver imo: if they lose everything, feel thrown out by society, what still drives them to do good afterwards? I'm sure many of them will simply want to better their lives but external motivation must also help

So I see it like the people who I saw saying upthread that rehabilitation should be the goal: if they're a danger to society, idk, whatcha gonna do but control that? Need to lock them up or similar (ankle thingy, idk). But if there's a good chance they'll get back on their feet and become taxpayers instead of prisoners or crime group members, then that's what we should asap strive for