I know the US is not the world but since we are discussing Iron Yard, it's worth pointing out that the private sector in the US is almost exclusively at-will employment.

Wrongful termination suits are something that all large companies have lots of scar tissue around though

Companies that size should have well-documented processes to follow to minimize liability around that. The much larger problem I've seen is that managers - since ground-level engineering managers tend to overwhelmingly be young and first-time-managers since the industry is dominated by youth overall - are green, don't know what they don't know about firing, and wait to bring in HR until they've already made up their mind, at which point they find out about the six month official process they should've started following six months ago. Or are just afraid of having the conversation, period.

New manager training is an even bigger hole in most orgs than on-the-job entry-level engineer training.

Small companies usually have a lot more leeway here, I've been the replacement employee for a guy who only lasted a month and a half at a startup before.

After getting past my initial uneasiness around dealing with poor performers (defining poor performer as "someone the rest of the team feels is a net negative and can't trust with important features"), I'm less afraid of a not-good-enough hire (effort and conscientiousness in looking for the right solution is the most important thing I look for there), and withhold my "definitely not" calls for people I think would be truly toxic to the team's rhythm.

Well that's because they have poor hr its not hard to fire a genuine NCI (non culpable incompetence) employee.

This is probably true. It's just been my experience that it is extremely hard to get managers to evaluate performance and reliably go through the long drawn out process of letting people know they are under performing, and then removing them.

That's not the whole explanation, IMO. If you had the same distribution of accepted candidates with both the loose and strict policies, I'd still expect interviewers to use the strict one.

Why? If their hire doesn't work out, they've got a much better story - "Our hiring processes are really strict and everyone got on board with it, it must've been something else that went wrong". And when the hire does work, people at the company get to self-congratulate about how "elite" they are.

The costs of extraneous interviewing and leaving positions unfilled is borne more by the company, while the costs of making a bad hire is borne more by the hiring manager, so the hiring manager makes choices that offload costs more onto the company. Principle-Agent problem.