Imagine IntelliJ and Eclipse are both movie critics, and we're competing to see who can be the best critic. IntelliJ uses software which lets them discuss using in-line YouTube embeds, so they can easily call attention to a scene or vocal inflection. Eclipse uses text-only mailing lists, requiring that discussion points be explainable in common prose. In this scenario, it is likely that the IntelliJ community would develop a sophisticated sense of color and sound, while the eclipse community would be unable to discuss anything besides story structure.
This is a forum thread
where the Atom developers are discussing coffeescript syntax. Although the discussion is about syntax, you'll see that it naturally segues into "how will we do highlighting?", and they toss around some visual mockups. They can easily and naturally discuss visual concepts.
Here's a VSCode bug report
on GitHub. The UI is jumpy, so the user just uploaded a gif of the jumpy UI, right inline with their text. How many sentences of monospaced prose in Bugzilla would be required to communicate what this gif communicates?
There has been some good discussion on whether or not we need more cathedral, more bazaar, or just plain-old more people. But I think it's important to remember how important tools are in all these scenarios. Beautiful artwork is valuable in cathedral-based and bazaar-based ecosystems, because both architects and the common person have eyes that can see. People come to bazaars and cathedrals because they can tell where to come in. Imo, we're made our devs blind and hidden them in a maze.
If you started an IDE project from scratch today, you could have forums on discourse and issue tracking, code review, and CLA on github in one day. Obviously it's not so easy for eclipse to make this change. But is the history an asset or a liability? Die an innovator or live long enough to see yourself become legacy.