The type of language change the students are shown undergoing would require more than a source of new lexical items, I would think.
We find morphological change: Wouldsest for 2nd person singular present of "would".
And syntactic change: What speechitated Harvard? for "What did Harvard say?" (note the necessity of do-periphrasis in modern English).
How could a thesaurus (of fake synonyms) drive these sorts of changes? [Of course, under Minimalism, parametric variation, including differences in word order, is theorised to be a reflex of formal features which are borne by lexical items. So perhaps if the thesaurus had some way of encoding abstract syntactic features in such a way that they would be picked up along with the phonological and semantic aspects of the lexical item....]