Open-source software systems have proliferated over the past few decades, with increasing penetration across domains. The wide availability of development data from such systems has led to studies on various aspects of the software development life cycle. A large majority of these studies conduct correlational analysis of historical data to identify associations between variables of interest. Due to their correlational nature, such studies cannot isolate causal effects on key software engineering outcomes. In this paper, we investigate how the uniformity of developer interaction affects bug resolution time. In large and complex ecosystems, developer interaction is influenced by the experience levels of participating developers. Accordingly, we classify developers on their co-commenting experience levels and identify different types of interactions across levels. Using the causal inference technique, inverse probability of treatment weighting (IPTW), and real-world data from two large-scale software development ecosystems, we establish a causal relationship between the distribution of developer co-commenting activities across multiple categories and bug resolution times. We find consistent evidence across both datasets that increasing uniformity of interaction distribution leads to increasing bug resolution time up to a limit, beyond which the effect is reversed. Our results can inform team assembly and governance, and guide individual developers in their choice of interactions with peers.