They were needed at some point. They did corrections and even peer reviews. Now it is just a lucrative enterprise.

But they sill serve as reference for how many papers someone published and how often someone gets referenced. The reputation of the journal is relevant as well. I think this worsens the replication crisis.