Elisabeth:
Sorry to say this, but in my opinion we shouldn't accept any new projects at all for the next two years but instead work on improving the existing ones. There's enough work to do and even the time we discuss new projects would be better invested in discussing and solving our current problems.
A blanket injunction against new projects may express popular sentiments, but is unhelpful. However, I would certainly agree that technical evaluation needs to take higher precedence for future projects, and where key technical requirements are not met, new projects should not be launched. Furthermore, in our own allocation of funds to development work, we should strongly prioritize existing projects, unless we receive a grant specifically to develop a new one. This, combined with the requirement that new projects shouldn't be too narrow in scope (Wikispecies is an unfortunate exception in this regard) should be sufficient safeguards against mindless expansion. Artificial barriers are not needed.
It's also important to note that ever since Wikispecies, we have become *much* more conservative with launching new projects. I find it highly unlikely that the majority of the community would support a poorly planned and insufficiently implemented project at this point. Nevertheless, open discussion of even the wackiest ideas should be tolerated. After all, Wikipedia started out as a wacky idea.
Erik