Data visualization is covered in academic education. Just check the online courses from Coursera and other MOOC platforms. The problem is data visualization is not enough – one needs to know about the business, about the problems the business is confronted with, respectively the data, processes and the logic that applies on top. That’s the gap between theory and practice.
Many dashboards are poorly designed and probably there are many practices that are not adequately enforced though the most wonderfully designed dashboard won’t help you if it doesn’t model the business problems from various perspectives. Being able to identify the most important factors has to do with the hands-on experience.
Just dumping some data in nicely designed visuals might be called data visualization though it barely scratches the surface. Occasionally, it’s what users need, or think they need. The problems of data visualizations are multiple – the lack of data culture, the lack of understanding the business processes and the data, the lack of being able to define and model problems, the lack of educating the users, the lack of managing the expectations, etc.