Tolya: To say it is a failure you would need to take into account how much money was spent on creating/implementing a DRM system. And establish that a DRM free game is more likely to be pirated. Then see if the surplus of illegal downloads offsets the costs of DRM.
Show me those numbers.
ET3D: I don't think all this is needed. First there's need to show that DRM-free games are not pirated more than DRM-laden games. If that's true, then GOG's stance is a success. What we want to prove is that DRM doesn't help, or, to put it another way, that DRM-free doesn't encourage more piracy. If that's true, then there's no point in DRM.
If DRM-free games
are pirated more, then arguing the cost of DRM would be quite meaningless. Sure, it would be an interesting intellectual exercise, but it won't really matter to the bean counters.
Except you can't measure any of it, because no two games are the same, no two same games sell the same. Therefore it is impossible to throw a number of alleged illegal downloads and say that something/someone has failed.