I mean, in this situation a big thing is taking a plugin and re-writing it as part of the core application. The initial plugin was inherently limited by the existing plugin API, which wasn't optimized for this specific use case.

"We improved a popular plugin by integrating it with our core application" doesn't feel like a confession to me. In fact, that's exactly what I'd like to see from an application developer. No application is perfect, and it's good when developers improve the basic look and feel of an application based on what their users want.

What if after a long research, you can change the complexity of an algorithm from O(N) to O(log N), which makes the code orders of magnitude faster on large datasets. Does it mean the original code was not "sensible code"?

I know reading the article is hard, but if you did you'd realize that it was slow because colorization was performed by an extension, and could not be optimized sufficiently because of limitations of the public API. They sped it up by moving it into the VS code core, which allowed it to take advantage of a bunch of features not available to extensions.

You didn't even bother reading the article, did you? The performance increase was achieved by inlining a third-party extension into VSCode itself.

The code was both mature and sensible _within the limits of the environment and API available to it_.

The VSCode team took it out of that environment completely, which opened up a huge number of optimization possibilities.