There are two very different questions here, though. One is whether we should accurately assess risk and probable healthcare costs. To which: yes, obviously.

There are two very different questions here, though. One is whether we should accurately assess risk and probable healthcare costs. To which: yes, obviously. But the other is whether our existing system can do that without ruining people's lives for reasons they can't control. To which: no, probably not. The article points out that for insurers (or employers who underwrite their own insurance), mandates to cover people mean that dropping/firing someone in a high risk pool makes more sense than pricing accordingly. We already have regular problems with people getting dropped as they claim, we should expect to see people dropped because they will claim too. Sure, risk pooling makes sense. But in practice we have a strikingly inefficient market that will probably respond to this information in a way that ruins a lot of lives unnecessarily.