For flows with hundreds of datasets, it'd be helpful to be able to visualize them more clearly. It'd be nice to be able to minimize portions of the flow into black boxes which can be opened to examine internal components. Often, many datasets are compressed down to just one dataset a few steps later, so hiding these would be ideal.
A workaround I tried was to create a hidden tag, then create a filter to hide everything with a hidden tag, but since these datasets had other tags, I couldn't get that to work. It'd be great if there were a built-in functionality to condense large flows.