Totally against using LeetCode for interviewing engineers. However, when asking around about why people who use it do so, I found out it does have one irrefuta

Totally against using LeetCode for interviewing engineers. However, when asking around about why people who use it do so, I found out it does have one irrefutable advantage: it stops people who can't code at all. From an engineer's point of view, LeetCode is a complete waste of everyone's time because it measures things that aren't factors in successful engineering (as TFA says). Bu from the non-technical manager's point of view it's awesome because it gives a single, simple score for "how good is this engineer compared to the other ones?" and people who can't code at all can't complete it. The non-technical manager's worry when interviewing is that they hire a really expensive employee who can't do the job. But because they don't understand the tech, and the tech is complicated and even expert engineers spend lots of time fighting it to no apparent end, it's really hard to understand if an engineer is incompetent and bullshitting them, or actually good but the problem is hard. Having a nice, easy metric that stops the complete bullshitters from getting in solves a problem for them. What we need, obviously, is a professional association for software dev, that can then properly test us and verify that we can do the things we say we can do. But the industry has a lot of growing yet to do to get to a point where this is even possible.