Correlation does not equal causation is one of the most widely misunderstood sciencey terms.
It would be better phrased as "correlation does not prove causation, but it does imply a relationship"
Yes just because something is correlated does not necessarily mean they the have a direct causal relationship, and more importantly it doesn't indicate which way a possible causal relationship would exist. But it does imply that they have a relationship, either directly or that they are both driven by a third force.
You have to apply some additional logic.
What's more likely?
A) Coffee consumption causes good health.
B) Good health causes coffee consumption.
C) Some other force causes improves health AND increased coffee consumption.
D) This is a spurious correlation.
B seems absurd. Or at least would have limited measurably effect. Yes healthy people are more able to drink coffee but it seems like such an effect would be limited.
C could be the case, but with coffee consumption being a deliberate action, if you believe in free will, a seems more likely.
Yes it could also be a spurious correlation, I guess that's possible too.
In this specific case, at the very least you can conclude that coffee is not acutely harmful. If coffee _was_ acutely harmful you would expect to see an inverse correlation between coffee consumption and general health.