The problem with these sort of analyses, in my mind, is that they seem to neglect the iceberg of code you build on top of, which works perfectly fine and is virtually bug-free. For instance, imagine that I'm building a website. If size is a predictor of quality, then you'd expect me to add bugs as time goes on, and that's probably true. But my site is stacked on top of Chrome, an abstraction with hundreds (?) of millions of lines of code that virtually never give me an issue. This is stacked on top of the typescript compiler, the C++ compiler, the operating system, etc, and those all run in the millions (TS) to tens or hundreds of millions (OS) of lines of code! So while I don't doubt that codebases get buggier over time, this seems to neglect the fact that I use truly immense amounts of code every day that don't seem to run into an issue. Perhaps the bugs grow around the edges rather than in the center, or something like that?