This feels very ivory tower. It reminds me of the "You should never need to check user agent in JavaScript because you should just feature detect!!". Well in the real world that doesn't work every time.
The same is true for server side applications of user-agent. There are plenty of non-privacy-invading reasons to need an accurate picture of what user agent is visiting.
And a lot of those applications that need it are legacy. Updating them to support these 6 new headers will be a pain.
Chrome will support the legacy apps by maintaining a static-user agent. It just won't be updated when chrome updates. If you want to build NEW functionality that where you need to test support for new browsers, you do that via feature detection.
The same is true for server side applications of user-agent. There are plenty of non-privacy-invading reasons to need an accurate picture of what user agent is visiting.
And a lot of those applications that need it are legacy. Updating them to support these 6 new headers will be a pain.