I feel silly responding to a downvoted comment, but here goes anyway: Correlation does not imply causation, but it wiggles its eyebrows suggestively and points “look over there!” (Thanks Xkcd)

In this case, we are in a position of not knowing very much about what causes/prevents COVID. Based on this study, we can learn Vitamin D seems to be involved, and there is an obvious causal mechanism that one can hypothesize might exist, even if not proven. Even if there’s only a 10% chance that vitamin D supplementation could be helpful, knowing it as a possibility could be useful to many, and justify action even if uncertain about the causality.