Visualization can directly engage the decision making part of your brain, that is, good visualizations can. So from that perspective, a good visualization is essential. A picture that communicates a concept in data is very valuable, especially to the extent that it helps you think about that data. Consider people trying to visualize cancer/health related data on their genome. Think there's a market for that? Giving the average person a vcf file is not sufficient - yet it can be a bioinformatics feat to produce. Give them an interactive interface full of visualizations, and you make a fortune and help human health at the same time :)
Especially "nowadays" as you say, when we have to deal with massive amounts of data across many dimensions. It's often a visualization that can bridge the gap and give people a handle for thinking about their data. So, one can do plenty of bioinformatics without visualization. But it can also be seen as a limiting component in many ways - thus it is a marketable and effective skill set to have or develop.
We all know people who make terrible, boring, useless plots. Hopefully we also know people that make insightful, informative, transformative plots. It's a skill set that needs attention and development. Anyone who's read Edward Tufte's books can site plenty of examples of plots/visualizations that, had they been better, could have avoided dire consequences.