1. are my false positives low?
2. is the time it takes to fill a position low?
If the answer to both of those questions is "yes" then the system is great for me (and I've seen the system described above provide affirmative answers to both these questions in a startup setting).
Determining whether a hiring system is locally good is significantly easier than determining if it's good in general.
The problem is that you don't know the cost of the false negatives; you don't know whether the people you didn't hire would have greatly improved your processes / code / business
It’s probably not the stuff you know how to do that makes or breaks your company. Either someone else does it, or you “figure it out” and hopefully you know if you did it right before it’s too late.
How do you hire someone to be good at stuff you aren’t already good at? There’s gonna be a lot of those in the false negatives pile.
Short of industry wide collaboration with some kind of longitudinal study, getting an idea of what does and doesn't work in terms of false negatives is nearly impossible.