Licensing debates abound in the "free culture" academic discussion, but we believe much of it suffers from one of two problems. First, restricting themselves to a purely technical, legalistic approach; this kind of approach is important and necessary for improving licenses, but it is not sufficient to deal with the social impacts and ramifications of such licenses. Second, adopting a dualistic and overly antagonistic approach, particularly with regard to the debate on commercial usages; this debate touches important aspects of the licensing discussions, but it frequently stalemates (due to its polarized character) before it is possible to make significant advances in deepening our knowledge about these issues, and coordinating efforts for improving the knowledge commons. The copyfarleft licensing model is a concrete proposal related to the commercial usages debate. It suggests one way to improve on "non-commercial clauses", arguably making them more effective in reducing negative social impacts, such as wage labour being exploited in building the commons. We believe it is an interesting contribution that mostly avoids the problems mentioned above, but also has its potential drawbacks. This position paper comments on them (as well as on some merits of this model), and proposes an alternative or complementary model that attempts to solve some of those drawbacks.