On a generalized framework for turbulent collision frequency models in flotation: the road from past inconsistencies to a concise algebraic expression for fine particles