Units Don't Lie

Two variables, height (in cm) and weight (in kg), have a covariance of 150. A colleague claims this means height and weight are strongly related. What is the main issue with interpreting covariance this way?

Covariance requires the variables to be independent to be meaningful

Covariance is scale-dependent, so its magnitude alone cannot indicate the strength of a relationship; correlation (which is normalized) is needed for that

Covariance can only be computed for variables measured in the same units

A positive covariance always means the variables are causally linked