The author was confused by the term "long tail". IMO this term is best used to visualize how a model works (Amazon's long tail of books, Google's long tail of niche keywords and search results...), but when applying this concept strictly to a specific model to see if it's a long tail business it can yield biased conclusions. He cannot find one specific album on the store doesn't mean iTunes is a no "long tail" store. Or like when I cannot find a book on amazon it doesn't mean Amazon is a no long tail business. So rather trying to conclude if iTunes is a long tail business by looking for one specific item and see if it's available, a better and more precise way to do it is to look at the stats (quantity of "niche" songs sold compared to the more popular ones, revenue earned from those niche songs ...) to see if iTunes selling a large number of unique items in relatively small quantities
Good point.Although the lack of specific items that specific consumers may well be interested in purchasing does indicate a failure to capture all of the tail. Other suppliers could potentially pursue said niche.
The article assumes that older releases are unavailable due to laziness. This is not always the case.
Recordings can be deleted for any number of reasons including: infighting, copyright issues, legal issues, embarrassment, ego, lack of interest, changing trends.. you name it.
In some cases the rights, once sold to a label, can exchange hands so many times, that ownership can't even be determined.
In particular, if any of the affected parties go out of business, the songs can end up in limbo for eternity.