Two structures or two sequence can either be homologous or not. Homologous means deriving from the same structure from a common ancestor. Bat wings and human arms, for example, are superficially quite different, but they are homologous structures. This fact become less surprising when inspecting the bones anatomy.
So… it makes no sense to talk about the percentage of homology, while it is correct to measure the percentage of identity of two sequences (that can suggest their homology).
I think this distinction is quite important, yet even very popular papers can help carrying this error:
