here's a trivial example of why the "standard" geometric mean  is not a good estimate of central tendency when you have values near  zero.

Data that combine multiplicatively, like rates, are actually very common outside of economics too. The key is to recognize when a measured variables is affected by many (semi) independent forces, each of which scales that variable up or down — rather than simply adding or subtracting a fixed amount to it. This is often true in the natural sciences.