No, the censorship was not at all limited to "stopping direct harm". Unless you redefine harm as "what Twitter owners/management decide is harmful", in which case that's true by-definition.

Classical case in point: https://twitter.com/JulianAssange