The conversation around this topic always seems directly or indirectly framed around a zero-sum framing: what's better and what's worse? Which side wins? Even if you disagree with the premise, that's what's shaping the conversation. I don't think the article suggested there's a "correct" answer, but it was clearly inspired by people who think the author was doing things wrong.
It can simultaneously be true that there are successful long-term campaigns with and without high character turnover due to death. It's a mater of personal preference and successful execution. The only thing categorically false is the idea that character deaths, in and of themselves, are inherently bad for long-term play.