The claim about Chaga is false w.r.t. "the only way" – it probably follows from a misstatement of a theory-internal claim that Chaga has no prepositions (Bresnan & Moshi) – depends on what you mean by "preposition". There is no way to compactly express "benefactive" vs. "malefactive", however the locative suffix is -inyi and the instrumental is a preposition na. There are constructions like "for the sake of X" that paraphrase the benefactive use of the suffix. Also, instrumentals may be treated as objects by addition of an affix on the verb and elimination of the preposition, but the affix varies within. The term "applicative" derives from the traditional Bantu term "applied" which is a specific suffix, *-ɪd- (il, ir), but the causative (*-ic- etc) is also used. I assume your question is about semantics and not Bantu-specific morphology. I don't know why they call it a "voice".
Anyhow, many Bantu languages allows promotion of multiple NPs to object status which triggers verbal affixation. Typically, this includes benefactives, locatives, causees, possessives – see Kimenyi A relational grammar of Kinyarwanda. There may be limitations on how many bare objects are allowed, thus you may not be able to get promotion of both a locative and a benefactive, or you may be able to. Some Bantu languages allow multiple applied suffixes for different functions. Here is an example from Kerewe. You can say bakategula empilya kuluga ha Bulemo "they took money (empilya) from Bulemo"; you can promote Bulemo and say bakategulila Bulemo empilya meaning the same thing. You can also use the applied for "because of", and then you will have two applied suffixes, thus bakategulilila omwaana Bulemo empilya "They took money from Bulemo because of the child (omwaana)". Double suffixing is not possible in Kuria, but it is possible in Kerewe – this is a basic difference defining one language vs. the other.