I don't know much about the Arabic script but Spanish got the grand total of 5 different vowels.
To transcript Portuguese would effectively be a much harder task. I wonder what mozarabic looked like in comparison to those two.
Well, that's only counting basic vowels without diacritics. Arabic would need to display these by creating new haraqaat, or vowel marks for all of these. Notice, there are already other types of diacritics used for supplements. This mind you would not include diphthongs, which the abjad would need to represent.
The same sort of issue occurred with Turkish which had too many vowels to really use the abjad. Abjads, while are the original alphabets, are not effective for non Semitic languages at the end of the day. I would suggest either you drop the abjad and simply use an alphabet with the Arabic script (with vowels as letters and diacritics for vowels not represented) or the language use Latin alphabet and then either use diacritics or spell out methods for Arabic inclusions.
So a question would be how does it represent, Semitic oddities such as ا the fall of Arabic L2 learners? You can use ā or how I typically write it transliterated, as aa.
There also should be a discussion as to the Arabic dual. Would this language adopt a dual in addition to the plural and singular? If so, that would be interesting. In this case, English would go best as a creole with Arabic as English has a sort of latent dual if I am not mistaken.