Pivotal Tracker. It's opinionated, but because it can rely on those assumptions, it strikes this incredible balance between information density, on-the-ground usability, and the ability to have a birds-eye-view of the risks of a software project. A diligently updated Pivotal project is a thing of beauty and confidence and inspiration. One size does not fit all (and certainly, if you want to give your team the opportunity to work on more difficult-to-spec, research-oriented, or experimentation-dependent projects, it's far from ideal), but for typical web applications, it's industry-defining.

Worst-designed software: Oracle PeopleSoft Campus Solutions - https://www.google.com/search?q=peoplesoft+campus+solutions&... . A case study in "how many workflows can we fit into one infinitely customizable software product without hiring a UX or UI designer to give any thought to a single one of them? Oh, and how can we make it as non-performant as possible to do the most common tasks?" I understand the siren call of "an interface that works for every workflow imaginable" - but unless you build design-customization tools on top of it, it will be a hated monstrosity.

Fun story: my undergraduate college used this, and I got a highly competitive internship after an interview where I spent the majority of the time ripping apart exactly how bad this software was, and how I would fix various aspects. I imagine no person motivated enough to solve those problems has any desire to work for Oracle, so the problem will persist for at least another century.