What he's saying seems to be this:
You start with a potential pool of sales - 100% of the iPhone user base. From that you want to compute "sales lost due to piracy". If you measure it by "only 20% of the people who play my game have paid for it, so 80% of my potential sales are lost due to piracy" then you get the measure that the industry usually talks about. But if you're talking "lost sales" then the article author is arguing that's not really right - if 80% of the people playing your game are pirates and only 10% of the population of customers who own the phone even have the possibility to BE pirates, then you've lost a lot more sales to some reason other than piracy.
I see where he's coming from - the stupid 80% number they bandy around is meaningless - but his reasoning isn't that much better. It's a different way to approach the problem, but he's missing the fact that the potential customer pool isn't really 100% of the people who own an iPhone, and that's the key question. If 50% of the people who own a phone are interested in your game that's one thing - the size of the potential customer base dwarfs the number of potential pirates. But if only 5% of the people who own the phone are interested at all in your game then the overlap between interested parties and potential pirates becomes much more important to your bottom line. And I suspect that when you see companies bandy around that 80% figure what that means is that the size of their base of people potentially interested in their game is small and has a high overlap with people willing to pirate it.
And that's not even getting into the behavior of pirates that's linked to in the article - where pirates are willing to download and try A LOT of stuff because it's free for them to do so, so they do it. But that doesn't mean they stick with it and use it forever. It's disposable to them - they play with it for a while and move on to some new shiny toy. Those aren't lost sales in any meaningful sense - if they hadn't been able to get it for free they never would have tried it in the first place. Which means that the overlap between "pirates" and "people interested in my game" is in many senses artificially high because the people in the group "pirates" are interested in almost EVERYTHING that they can get their hands on, even if only for a short period of time, because they're not paying for it.