This is years late, but offering it anyway:
Ethnologue tends to approach language from what I've heard called a "splitter" perspective, where in effect certain differences are considered sufficient to define separate languages. Where the names (endonyms) used for the languages are also different, that may reinforce the idea of separate languages. In Africa there are many cases of this.
The splitter perspective is not wrong, but it is also possible to work from a more "joiner" perspective, understanding that mutual intelligibility, even if not absolute, is sufficient to identify a larger language. Ethnologue (and the code space SIL manages, ISO 639-3) has a few "macrolanguages" that would be examples, however these were inherited from older coding systems. Ethnologue (nor SIL & ISO 630-3) has not sought to identify macrolanguages beyond that. There are no macrolanguages in Cameroon other than Fula (ff or ful), which of course is only in the north of the country as well as most of West Africa.
The Pan African Localisation project adopted the more joiner approach, informed by that used by some American specialists in African languages and linguistics (notably Prof. David Dwyer, Michigan State U., who compiled a Handbook of African Language Resources, and also the US National African Language Resource Center). One grouping of more or less intelligible languages largely in southern Cameroon goes under the name Beti, and includes among others, Ewondo, Bulu, and Fang.
While Beti is apparently not now a formal language category, and there's some history behind the retired code element for it (btb), it might guide in identifying some important languages that share similarities in structure and vocabulary that could be useful for translation and localization work. The total of all L1 & L2 speakers might be around 2 million. A page on the old PanAfrLoc wiki might be of use, although it needs updating: http://www.bisharat.net/wikidoc/pmwiki.php/PanAfrLoc/Beti