In view of the fact that Arabic omits the diacritics for short vowels and the like, does this pose an extra challenge for text to speech systems for the Arabic language? How easy is it to resolve these challenges?
No, lack of vowels or diacritic vowel marks is not a particular problem for text-to-speech systems, which usually make use of a dictionary of phonetic transcriptions anyway.
Very few languages have a so tight mapping between writing and speech that correct pronounciation can be deduced by its written representation alone. Without a phonetic transcription dictionary, even English would be nearly impossible to pronounce. Just compare words like earth, hear, heart and break and then try to find a rule for how 'ea' is supposed to be pronounced.