In 2016, a substantial update was made to this database. All transcripts not labeled mRNA (e.g. non-coding, micro, etc.) were removed. Genes from the ensembl database were added. A tweak was made in the method of identifying redundancy that removed more transcripts. The new database contains 27,012 transcripts.
The new database (with largely old names intact) is here.
A table that attempts to reconcile database names with current NCBI names as well as ensembl names can be found here.