And even then it’s probably not a hard rule as much as a good heuristic: the older a source is, the more careful you should be citing it as an example of current understanding, especially in a discipline with a lot of ongoing research.

If somebody did good analysis, but had incomplete data years ago, you can extend it with better data today. Maybe the ways some people in a discipline in the past can shed light on current debates. There are definitely potential reasons to cite older materials that generalize well to many subjects.