Spatial audio representations have been shown to positively impact user experience in traditional, non-immersive communication media. While spatial audio also contributes to presence in single-user immersive VR, its impact in virtual communication scenarios has not yet been fully understood. This work aims to further investigate which communication scenarios benefit from spatial audio representations. We present a study in which pairs of interlocutors undertake a collaborative task in an audiovisual Virtual Environment (VE) under different auralization and scene arrangement conditions. The novel task is designed to encourage simultaneous conversation and movement, with the aim of increasing the relevance of spatial hearing. Results are obtained through questionnaires measuring social presence and plausibility, as well as through conversational and behavioral analysis. Although participants are shown to favor binaural auralization over diotic audio in a direct active-listening comparison, no significant differences in social presence, plausibility, or communication behavior could be found. Our results suggest that spatial audio may not affect user experience in dyadic communication scenarios where spatial auditory information is not directly relevant to the considered task.