The main logical failure of this article is the wrong assumption that the code published some time ago does always reflect current skills or relevant to the cur

The main logical failure of this article is the wrong assumption that the code published some time ago does always reflect current skills or relevant to the current needs. What you have written even couple years ago is quite often no longer important: progress is fast, new design patterns, new libraries, new, more advanced coding approaches become popular, you learn, you continuously improve your skills only to be judged by what?.. Sometimes, in some cases, it makes sense to filter candidates by looking at their GH. But I would not bet on that: the better filter is the code being written today, which is unlikely to be available for most of the developers.