As far as I remember, in the early days it was assumed that most spread was via visible droplets and surface because all well-studied viruses that spread by aer

As far as I remember, in the early days it was assumed that most spread was via visible droplets and surface because all well-studied viruses that spread by aerosol would spread much faster in a zero-immunity population than observed with SARS-COV2. But those are all viruses that have many generations of adopting to humans under their belt. Compared to them the new one seems to be barely able to make landfall in humans at all. Now we know full well that the virus is aerosol-capable and the implied "surprisingly bad compatibility with humans" is consistent with the number of observed mutations that spread faster, probably quite a few evolutionary low hanging fruits to increase hit rate while spreading. Back to topic, I think that in early days the surface focus wasn't a bad guess, it just happened to be wrong.