The general point here is that single plane models of context tend to introduce additional complexity to handle the contextual variables distributed among field, tenor, mode and genre by English Text. So the cost of recognising two connotative semiotics instead of one is not as high as it might initially appear.
This is the conclusion of Martin's "argument" for the theoretical superiority of his stratification hierarchy over Halliday's original model, and it is clearly the opposite of what is true.
 As the previous six posts have demonstrated, Martin's model is more complex, not less complex than Halliday's model, but more importantly, its complexity arises from misunderstandings of the theoretical architecture of SFL, principally stratification and instantiation. This has resulted in a model that is inconsistent both with the theoretical architecture and also with itself.
Furthermore, in the course of his argument, Martin has been prepared to misrepresent the work of both Halliday and Hasan in order to invent non-existent problems, which he purports to solve with his model. As the previous posts demonstrate, each of Martin's "solutions" involves the creation of additional theoretical consistencies.
 The principal defect in Martin's bi-stratal model of context is, of course, that it is not a model of context. Register and genre refer to language, not to context, and in SFL theory, they are complementary perspectives on diatypic variation.
With respect to register, Martin's misunderstanding arises from mistaking the contextual systems that identify registers — field, tenor and mode — for systems of registers themselves. This in turn arises from Martin not understanding the principle of stratification: as levels of symbolic abstraction related by realisation (as demonstrated many times in previous critiques).
With respect to genre, Martin's misunderstanding arises from mistaking Hasan's generic structure potentials for potentials 'at the level of genre'. For Hasan, they are structural potentials at the level of semantics which vary according to genre (text type/register).
In conclusion, because Martin confuses semogenesis (all strata make meaning) with stratification, his strata of register and genre are — ignoring internal inconsistencies — both focused on linguistic meaning; that is, both strata are theoretical relocations and rebrandings of semantics.