Compounding is very rare in Semitic, which appears to contradict the claim.
The following is from Orin D. Gensler, 'Morphological Typology of Semitic', in Stefan Weninger (ed.), The Semitic Languages: An International Handbook, pp. 287–288:
Compounding as such is almost unknown in Semitic. There are of course lexicalized or semi-lexicalized collocational combinations of two nouns, but structurally these are normally indistinguishable from the Head-Genitive syntagm of the Construct (see 5.3.), and the two collocated nouns may be interrupted by the definite article (just as with the Construct). Thus in Arabic we have ibn al-sabīl (lit.) ‘son of the road’, i.e. traveler’, with the article al- separating the two combined elements. Only rarely do we find true compounds like Arabic rās-māl-ī (lit.) ‘head-capital-Adj’, i.e. ‘capitalist’. In modern (not Biblical) Hebrew, new compounds are not infrequently coined by blending two roots: for example, from midraḵa ‘sidewalk’ + reḥov ‘street’ the language has created the compound blend midreḥov ‘pedestrian mall’ (lit. sidewalk-street). In the modern Ethiosemitic languages ‘compound verb tenses’ are formed, some of which involve the compounding of a verb form and an auxiliary verb, notably ‘to be’; thus in Amharic təsäbr-alläčč = ‘she will break’ (lit. she.will.break + she.is)