In this commentary, we argue that studies similar to Cennamo, Marchesi, and Meyer (2020) should distinguish four dimensions of control in blockchain governance, all of which could be more or less decentralized. For some of these dimensions, decentralization is likely beneficial, while for others centralization may be preferable. Cennamo et al. (2020) provide evidence that the initial design stages of blockchain systems benefit from centralization. Future research is needed to provide empirical insights for other dimensions of control in blockchain governance.