For the most part we don't turn off flags in this way; at most we unkill flagged stories so ongoing discussion can continue. The flagging system works well overall.
But the community is divided on this issue. Many users flag these stories, but many other users care deeply about them. Those users are also a valuable part of HN. They argue that if we never turn off the flags then we are de facto taking this issue off the table (or allowing a segment of the community to do so). That may be a bit exaggerated—some stories about sexism and/or gender in tech do stay on the front page despite flags—but it's still a fair point. Whatever you think about these questions, they're not off-topic for HN the way most politics are. Moreover, many HN users who make this case are active members who participate in lots of threads. They're not just trying to use HN to further an agenda.
If we had better software, maybe it would all balance out in a way that didn't require any intervention, but that's not where we're at (and in fairness is a pretty hard problem). In cases where the software leads to unrepresentative outcomes, it's our job to adjust. Is this article the best one for that? I don't know, but given that the thread turned out to be a vigorous discussion that managed not to be horrible, I think it was an ok call.
But the community is divided on this issue. Many users flag these stories, but many other users care deeply about them. Those users are also a valuable part of HN. They argue that if we never turn off the flags then we are de facto taking this issue off the table (or allowing a segment of the community to do so). That may be a bit exaggerated—some stories about sexism and/or gender in tech do stay on the front page despite flags—but it's still a fair point. Whatever you think about these questions, they're not off-topic for HN the way most politics are. Moreover, many HN users who make this case are active members who participate in lots of threads. They're not just trying to use HN to further an agenda.
If we had better software, maybe it would all balance out in a way that didn't require any intervention, but that's not where we're at (and in fairness is a pretty hard problem). In cases where the software leads to unrepresentative outcomes, it's our job to adjust. Is this article the best one for that? I don't know, but given that the thread turned out to be a vigorous discussion that managed not to be horrible, I think it was an ok call.