I actually think it's a relatively interesting measure, they've just pulled some completely stupid conclusions from it and there's no way you could prove that there would be any regression to the norm from the stats.
If a player scores more goals than expected from the chances created for him it shows that they had a better season than could realistically be expected. For the same reason I'd be interested to see how well, or badly, our strikers last season performed on that basis.
All it shows is that last season Kane had a good season, something we already knew, and that the author has no idea what to do with his data so wrote some rubbish.