It’s so easy to bash interview processes. You can look at *any* process and describe the types of (good) candidates it fails for.
This post just criticized existing process without offering an alternative. Well, sure, it has problems!
However, the problem the author describe are not entirely accurate.
The post seemed to be describing two different interview approaches, without really making the distinction clear. It described algorithm/problem-solving interviews, as well as knowledge-based interviews. The criticisms of one do not really apply to the other.