An interesting question to ask - well, it's several questions, actually - is why and how 'grammar' staged a coup and took control of 'good writing' in primary schools?
1. First of all, it's only one kind of 'grammar'. Please, please, please don't be kidded into thinking that the stuff the DfE are dishing up as 'grammar' IS grammar. Grammar is the means by which we string words and noises and symbols together to make sense. This grammar is there in place doing its work, whether we have names for it or not. Then there is terminology to describe that grammar. The moment we come up with terminology, we can ask of it, is it describing grammar properly and what are the principles that lie behind the terminology? A good deal of the terminology is what I would call 'self-serving'. It treats language as a 'sealed system' with little or no relation to meaning and function or purpose in our interpersonal relations. So, if I say, something is a 'conjunction' - all that tells us is that it con-joins two things. It doesn't tell me anything about the resultant meaning or why I would want to conjoin anything. Yet, if I say a word is a 'possessive' - this is a completely different kind of description. It says that I, you, he, she or they 'possess' something. This is then connected to meaning and at least some of its function. So the terminology in traditional grammar is not even consistent with itself! Some of it pretends to take no notice of meaning and actually takes no notice of function. Other terms like 'possessive' take some notice of meaning and function. However, if you think about it, even self-serving terms like 'conjunction' only pretend to take no notice of meaning, because how would you know it was 'con-joining' if you didn't understand the chunks of language it was con-joining?! It just that it doesn't acknowledge this in the term itself.
2. If you look at the grammar foisted now onto schools, teachers, parents and children aged 5-11, it is mostly of the 'self-serving' kind. Teachers are forced to stand in front of classes and tell them about the virtues of 'expanded noun phrases' as examples of 'good writing', along with, say 'fronted adverbials'. To be clear - this is complete garbage. There are no intrinsic reasons why an expanded noun phrase is 'better' than a non-expanded phrase, or a fronted adverbial being better than a 'rear' adverbial. There can only ever be reasons why one is 'better' than the other, if the meaning and purpose are right - or, in short, the context. Sometimes it's good for effect to string a load of adjectives and adjectival phrases, and adjectives modified by adverbs in front of a noun. Sometimes it isn't. It's cods to say otherwise. Cods rules. Same goes for the frontal adverbials.
3. What has happened here is that one specific kind of knowledge (I'd call it 'false knowledge') about language has managed to declare itself THE knowledge about language and that this particular kind of knowledge is what helps you write better. Both these statements are false. There are many kinds of knowledge about language. There are many branches of linguistics and applied linguistics. This false naming of parts approach - with its origins in written (not spoken) Latin, is one kind only. More importantly, if we start with the proposition that we all want children to write well, then we should then ask the question: what kind of knowledge about language will most help to produce good writing? I would suggest that it's the kind of knowledge that comes from investigating language, learning from language, investigating (for example) how my speech might differ from yours, from my parents, or, importantly, from some writing that I have in front of me. It's the kind of knowledge that comes from people who have tried to describe language in terms of meaning and purpose - which means identifying 'semantic fields' and 'intention fields' and the like. So, with exclamations (as discussed on a previous post) the issues to help with writing is 'how do we exclaim?' We exclaim in many different ways. What are these ways? Let's look at some exclamations in novels, newspapers and online. Are any ways better or more useful than other ways? When might we use one way rather than another? We might look at some of the terms to describe these and, indeed, just like any scientist would, ask ourselves whether these terms are good enough or useful enough. Meanwhile, the simplest and easiest way of investigating this with young children is looking at their favourite writers, see how they do things and say, 'Well, we could do that, couldn't we?' and we can imitate, adapt - and importantly, invent new ways derived from the ones we're looking at...or ones that just come into our heads. (That's more or less how I run poetry workshops.)
4. How did this stuff stage the coup? Because the 'naming of parts' approach can be marked as right or wrong. That's all. Not because it's right or suitable or that it is better at producing good writing than other approaches to knowledge about language. Simply and only because it produces right and wrong answers. This both satisfies a never-ending wish to assess teachers (rather than help them), and a need that some people have that there is only one way to speak and write - the one that is laid down by some authority somewhere rather than our own democratic ability to talk and write to each other and control ourselves and how we want to do it. This is just too democratic for some people.