The spirit called Marbas in the Goetia is Barbas in Weyer. The two forms sit at the same place in the same catalogue with the same offices, and a single letter has moved.
This is normal. The names in this corpus have travelled through Latin manuscript, Latin print, English translation, English reprint and English manuscript copy, and at each stage a copyist could mishear, misread or improve. Minuscule b and m are not easily confused, but they are not the only pair that has shifted: the catalogue also gives Samigina for Weyer's Gamigin, Sallos for Saleos, Cimeies for Cimeries, Haures for Flauros, Naberius for Cerberus in some readings. A reader comparing editions will find dozens of these.
One later layer deserves particular caution. The 1904 edition prepared by Samuel Liddell MacGregor Mathers and published by Aleister Crowley is the version through which almost every modern reader meets the text, and it introduces a number of alternate spellings of the spirit names that Joseph Peterson, working from the manuscripts, reports he cannot find support for in any of them. Where a modern book gives an unusual form, the question worth asking is whether it comes from a manuscript or from 1904.
None of this means the names are meaningless. It means that arguments built on the shape of a name, the derivation of Bael from Baal, of Caim from Cain, of Astaroth from Astarte, are being run on evidence that has passed through several hands, in several languages, over at least a century, with no critical apparatus at any stage. The resemblance may be original. It may also have been introduced by a copyist who thought he recognised something.