I don’t think that is a fair argument in this day and age of software development, especially for an operating system. With that level of complexity, I would contend that it is next to impossible to identify potential failure scenarios. I also think this suffers from a rose colored glasses view on history. Perhaps software in the past was as vulnerable, it just never got patched because there wasn’t an easy method to apply updates. Now that there is, it is much better to have a responsive development team to react and fix obscure problems that are difficult or impossible to predict.