I feel like there's an analogy to software in this story. Someone put in a hack that, while ugly, works. I'm sure many city planners over the years have looked at it and thought it should be fixed. But when the current solution is working, taking the time to do the necessary testing to ensure that a new, cleaner solution works would be painful. So, 30 years later, the hack is still in place.

I know I've come across the software equivalent of this many times in my life as a developer. "Hmm...that's a really odd way of doing that, I should refactor this. But I need to write a bunch of tests for this since the original developer didn't write any. And am I really thinking through all the possible use cases properly? Even if I am, this is now a 3-day estimate, which means it will probably take 5 days, and all to "fix" something that's technically working. That's a lot of risk/time commitment for very little payout. Maybe I should just leave it be and move on to a project that's better setup to succeed."

Yep...makes total sense that this light still exists. Still doesn't explain why they replace the green and amber bulbs though...