The article is interesting food for thought, but there’s a fundamental error. In the toy algorithmic problems, having the wrong big O solution is wrong because it doesn’t finish in a satisfactory amount of time. From this he seems to infer that anything that takes longer than another alternative is “wrong.” But the thing that made the toy solution wrong does not transfer to the other example problems! An editor dropping frames clearly isn’t sufficient to render it unusable for a significant slice of the target audience (proof by existence).