Most academics would agree with them that dynamically typed languages are a misnomer and in fact have no type system. And whether you want to use the colloquial term that counts dynamically typed languages or the academic one that doesn't is a subjective decision, and one where their choice was clear from context.
More generally, advocates of dynamically typed languages seem to want their languages to be considered typed because there's a developing consensus that typed is better. Unfortunately for said advocates, that consensus was not developed in contrast to languages like Forth, it was developed in contrast to languages like Ruby, so changing the definition of a type system to accommodate Ruby doesn't actually get you credibility in the new "cool kids club". You're better off arguing why dynamic is better than static than you are trying to enforce a particular definition of types. No one is going to be persuaded that Ruby meets their needs just because you made them use your definition.
More generally, advocates of dynamically typed languages seem to want their languages to be considered typed because there's a developing consensus that typed is better. Unfortunately for said advocates, that consensus was not developed in contrast to languages like Forth, it was developed in contrast to languages like Ruby, so changing the definition of a type system to accommodate Ruby doesn't actually get you credibility in the new "cool kids club". You're better off arguing why dynamic is better than static than you are trying to enforce a particular definition of types. No one is going to be persuaded that Ruby meets their needs just because you made them use your definition.