most often used messages" are called that for a reason. They are the messages that most often used. As software evolves, the messages that are most often used change with the time.
At the Berlin Hackathon, Siebrand was provided with new statistics and he has changed the composition of this group.
When you look at the group statistics, you will see that as a result there are now only 51 languages who have a complete set of localised messages. I do not have to repeat why it makes sense to do these messages first, but hey, they are the ones who are most often in your face. So do check them out...
The language committee will be processing all the outstanding requests and, all the languages that qualify for their first project will be judged against the old most used messages. Obviously, from a usability point of view even these projects do benefit from at least completing these messages.