Here's my take on it all:
FH didn't really spend any time developing his future languages; he wasn't another Tolkien. Judging from the examples in the books, Fremen is pretty much just romanized Arabic with some loanwords from a lot of different languages; Chakobsa is Serbian or Romany (Gypsy) and Maker-knows-what-else; Ixian appears to be Chinese; Galach is virtually non-existent.
I suspect that for anything besides the Fremen/Arabic (where he obviously used a dictionary and that British travelogue and some books on Islam—he quotes from the hadith in Children
, I believe—and may have even consulted an Arabic-speaking informant or expert?), FH just grabbed a dictionary or grammar or other book off his shelf and picked a word or example sentence at random.
Since there is so little Chakobsa given in the books, we can't really use it to say anything about the origins of the language. (Unlike with the Fremen, where we have the language and the religion pointing to some Arabic/Muslim ancestry.)
My solution is just to say that the examples of Chakobsa we have look identical to modern-day Serbian, etc., but that this is just an accident of convergent evolution; an accident, in other words. If someone were to try to develop a Chakobsa conlang, I wouldn't recommend them basing the grammar or other features of the language on Serbian. Maybe grab a few more words from that language to keep the same "feel", but nothing more. And I would reanalyze the examples we have syntactically and semantically and assign them the meanings FH gives them in the books.