Other strategies such as word recognition and just remembering words that they come across rather than having to decode it every time make for a poorer reader?
As your original statement was this: help them develop other strategies for working out unknown words, I'm completely failing to see what 'word recognition', which implies the child already knows the word, has to do with this. Word recognition is the end result of sounding out and blending, not a 'strategy'. Likewise, 'remembering' the word is also an end result of sounding and blending not a strategy.
If so, why do we teach children the list of high frequency words, words which are not phonetically regular, alongside those that are?
Well. It's a long story.
It starts in the days when phonics was completely taboo and Look & Say reigned supreme. Suddenly fewer children were learning to read effectively. so someone came up with the bright idea that if children were intensively taught to remember the most frequent words which occurred in text (by memorising them as 'wholes', of course) they'd be able to read at least 50% of the texts they were required to read. Sadly this didn't really work very well either because most of these words were 'function' words (the, and, off, in, up, down etc) which only served to connect the more interesting, but less frquently occurring, content words. The ones which actually gave a clue as to what the text was about. However, the High Frequency Word myth was out of its bottle and there is no putting it back. Witness you, and at least one other person, perpetrating it in this thread. Children have got to learn the HFWs. Well, children really have to learn all the words they're likely to encounter in their reading journies. Along the way, as phonics began to creep back under the radar, it acquired an extra facet as it was declared that these HFWs were 'not decodable'. Well, if you've ever looked at a list of HFWs you will find that is palpable nonsense. There are a few which contain unusual or unique letter/sound correspondences and which are useful in early reading texts (though not essential) such as 'one' and 'two' but the rest are easily decodable as long as you know the code. The only reason the Jolly Phonics authors dubbed them 'tricky' was that they did contain very unusual code, or code that the children had yet been taught.
Strategies such as using the context of the sentence or paragraph as a whole to work out what the word could be and decide whether their chosen word 'fits' mean that child has impaired reading ability?
Of course it does. What if they decide on the wrong word; it makes sense, fits the context but isn't actually the word on the page? I'm afraid in my book, and in that of most common sensical folks, that is 'guessing' and guessing isn't reading. Child should be able to work out what a word says in or out of context.
I don't mean simply guessing what that word could be, by the way. I mean looking at the initial sound of the word, or other sounds within it, and the length and trying to work out what it could be and then using their understanding of the text to make sure that it 'works' within that sentence.
As opposed to knowing the letter/sounds correspondences and being able to sound out and blend all through the word in order to identify it? Sorry, your strategy sounds very tricky and labourious and disabling. Incidentally, I've worked with children who have supposedly learned this 'strategy'. Sadly, their default position is look at the first letter and guess. It's much easier than anything else.
but what happens when we come across a word that isn't phonetic
Well, for a start there are no words which aren't 'phonetic' , though you might consider this to be a technicality. Phonetics is the study of speech sounds; all words contain speech sounds so none are 'not phonetic'. I suspect you mean words which don't appear to be phonically regular in that the letters don't represent the sounds you would expect them to. Dealing with them is a product of good teaching, and understanding of how phonics works and a good oral and written vocabulary.
or if trying to decode it phonetically could lead to a number of possibilities, many of them wrong? Using contextual cues in those situations is incredibly helpful.
I agree. Though I think that 'many' is a bit of an exageration. You're most likely to encounter only 2 or 3 options. Yes, context is very useful in this instance if you already have the word in your oral vocabulary.