David Rose completed his third reply to Erich Steiner on sysfling on 16 May 2018 at 11:47 as follows:
One advantage of this model is that discourse systems can be described metafunctionally, and explained stratally in terms of social functions. So patterns of tenor tend to be realised by patterns of appraisals and exchanges, patterns of field by ideation and conjunction, and mode patterns by variations in periodicity and identification.
One implication of such a Hjelmselvian model is that systems at each connotative and denotative stratum are instantiated simultaneously as a text unfolds. That is, the inter-stratal relation is not only a hierarchy of abstraction, but also of co-instantiation (or coupling), with each stratum contributing different orders of meaning making resources as texts unfold.
 This is misleading. This is not an advantage that the discourse semantic model has over theory-consistent models of semantics in SFL, since this is merely a statement about the SFL hierarchy of stratification that all SFL models share.
 The term 'social' here is potentially misleading, since it is the culture as a semiotic system that is realised by the semiotic system of language, not the social system in the sense of the material order of sayers (and sensers) who project the semiotic order of culture realised by language.
 To be clear, in SFL theory, these are patterns of instantiation — feature selection — during logogenesis, with different choice frequencies in texts as instances of different choice probabilities of different registers.
 To be clear, Rose follows Martin in misconstruing tenor, field and mode as dimensions of register misconstrued as context. For some of the theoretical misunderstandings and inconsistencies in treating register as context, see here (register) or here (context).
 To be clear, 'exchange' and 'appraisal' are not "discourse" systems, but genuine semantic systems. At the beginning of his reply to Steiner, Rose identified discourse systems as the grammatical systems of cohesion that Martin reinterpreted as his own discourse semantic systems.
 To be clear, 'ideation' and 'conjunction' are Martin's misunderstandings and rebrandings of Halliday's cohesive systems of lexical cohesion and cohesive conjunction (textual metafunction) as experiential and logical systems. As textual systems, their contextual counterpart is mode, not field. For some of many theoretical misunderstandings and inconsistencies in Martin's 'ideation', see here; for some of the many misunderstandings and inconsistencies in Martin's 'conjunction', see here.
 To be clear, 'periodicity' is a reinterpretation and rebranding by Martin & Rose (2003/2007) of what Martin (1992: 392-3) construes as interaction patterns between strata, with strata misunderstood as modules of meaning. It is largely a confusion of writing pedagogy and linguistic theory, where pedagogical terms (introductory paragraph, topic sentence, paragraph summary, text summary) are rebranded as theoretical terms (macro-Theme, hyper-Theme, hyper-New, macro-New). For some of many theoretical misunderstandings and inconsistencies involved, see here.
 To be clear, 'identification' is Martin's (1992) misunderstanding and rebranding of Halliday's system of cohesive reference. It confuses, for example, reference with nominal group deixis, the experiential meaning of the referent with the textual means of referring, and the immanent perspective on meaning with the transcendent perspective. For some of many theoretical misunderstandings and inconsistencies involved, see here.
 This is misleading. This is not an implication of the discourse semantic model in particular, since it is merely a confused statement about the SFL dimensions of stratification and instantiation that all SFL models share.
 This is misleading because it conflates the dimension of stratification with the dimension of instantiation and misrepresents the confusion as a theoretical insight (Martin's "co-instantiation or coupling").
 With the wording 'different orders of meaning', Rose is repeating Martin's misunderstanding of all strata as modules of meaning, which arises from Martin's misinterpretation of 'all strata make meaning' (semogenesis) as 'all strata model meaning'. See, for example, here.