I think you've hinted to the point that the article failed to make.
Processes are great, and absolutely required. But if you spend more time on the process than on the work then you have a problem. All the pre-work and planning might save you from a costly refactor once you find out you went down the wrong path, but honestly I've seen projects that had all the tests written up front and months of architecture planning and sprints planned only to jump into development and find that the whole approach wasn't feasible. Had they jumped into coding sooner they could have saved themselves close to a million dollars. I've also seen the opposite, the planning paid off and they had a great product, but they were late to market and it never sold well. Their competitors made it to market faster, their product wasn't as good but it was there and it improved over time.
Also, processes aren't going to protect you from bad developers. It just allows the bad developers to burn through budgets longer before they prove that they are not up to the task.