[orisa2]

I'd also suggest that another factor is that the author assumed that his reader base would have basic statistical knowledge, which I'd suggest would not be true even among software engineers. Just erase any reference to distributions or averages and you'll see what most people see. This just adds to your points.