What a terrible article. I feel dumber having read it. It's one of those articles that sounds good and takes more effort to correct than it took to write, and thus causes a net loss.

Just one example: "That gives every path to this function a 50% to 75% chance it will fail with just the inputs above." I mean, what kind of nonsense is this? This is pretty much saying that the odds of winning the lottery are 50%, because either you win or you don't.

Most of the article is spent on discussing the perils of a strcpy reimplementation. Yet there's no discussion of a fix. The author simply assumes the existence of a magical "safercopy" function that behaves correctly on all inputs. Of course the author doesn't provide the implementation of this function, because if he tried to write it, he'd discover that it's impossible.

In short: flagged for being awful.