No, the cost of DRM do not repay itself by preventing enough users. There have been tests cases, studies, and the repeated conclusion I keep seeing is that DRM has no correlation to the revenue from individual music, games and films. It has however a correlation to platform and user retention, which is what this article continuously refers to.

To take a few examples, remember when the European Commission hired researches to do a study on the effect of piracy? What about the revenue from itunes after it went DRM free? Sale of music that is available on youtube after it went HTML5? Revenue from music, film, and games on the longer term after aggressive blocking of torrent sites? What about the long term effect from anti-piracy/DRM laws?

The whole EFF article ends on the conclusion that DRM will not exist if their only purpose is to prevents enough users from copying the product. That effect won't be worthwhile enough to pay the cost of DRM. Only a sticky platform that prevent competition will bring the revenue to validate investing into DRM.