I think the COG database is not regularly updated or maintained anymore. In our lab, we use the eggNOG database, which contains information from the COG database and updated clusters based on a larger number of genomes. The most recent publication for eggNOG came out last year in NAR, so its still pretty active.
Well from the Clusters of Orthologous Groups of proteins (COGs) website they published a paper in 2003 "The COG database: an updated version includes eukaryotes.", and looking at their FTP site there seems to have been some data released in 2012.
For a definitive answer I suspect your best bet is to contact the folks at NCBI as detailed at the bottom of the COGs web pages:
Comments and questions to firstname.lastname@example.org or to the project supervisor.