Have we ever done a survey to see how important dark theme is to users? How they stack it against responsiveness, ease-of-plugin development, etc? I'm sure some users say it's their #1 issue, but definitely not all of them.
It appears that Lars and Ian are some of the most talented committers we have. Are we sure their time wouldn't be better spent on other issues? Obviously, it's up to them to decide how to spend their time, not me. But I'd like to make a final plea :)
Facebook, GitHub, StackOverflow, Box, DropBox, iOS - none of these products support themes. Do some of their users want themes? Sure! But even they only have so many development resources. Once a platform supports a theme, everything that plugs into it also has to support themes, so it puts a burden on the entire ecosystem.
IntelliJ, SublimeText, VSCode - these all support themes. And they all do so by using a UI technology which is fundamentally themable, but also fundamentally slower than SWT. They get themes for free, but they have to work hard to get performance and accessibility. Eclipse still wins the keypress-responsiveness competition.
Software isn't always about tradeoffs - sometimes you can improve something to the point that it's just fundamentally better than its competitors. When the Eclipse CSS engine was conceived, it wasn't clear that it might not be easy to use Windows MFC to create a platform which is just as themable as Swing or JavaFX or Web. But we've run the experiment for 6 years, and it's turned out to be very expensive, and still incomplete. Anytime we say yes to something, we say no to every other thing we could have done with that time. I'm baffled by the priority that dark theme has gotten with our most capable committers - aren't there other usability issues which rank higher?