I agree that it makes sense to teach sounds in contexts of words. I don't think I claimed differently - just that it is possible (and, IMO, can be meaningful) to learn to discriminate sounds (in spoken words) without at the same time learning the symbols for the sounds.
Getting back to how that might translate into teaching 'reading music notation' it makes me wonder what the equivalent to a word would be, a 'musical word' so to say. Rather than just playing a b-flat note and then expecting that to have any 'meaning' to a learner (for in the next step to introduce the symbol for b-flat); would it make sense/be necessary to learn to hear the various notes in the context of them making up a 'jingle' (for want of better ideas)? With my very limited musicality, I am totally lost when asked to identify a single tone, but can easily discriminate two tones from each other when played subsequently. And could make a guess at intervals (if that is the correct term).
But then I'm just thinking aloud here - probably starting to contradict myself ;)
IIRC that particular book I mentioned earlier, recommends starting to learn to discriminate the sounds (in words - spoken words) for average 2-3 year olds - as I said, pre-reading skills. Some children that age will be interested in the symbols for the sounds, some won't; and there is plenty of time for the symbols yet for those who aren't (IMO). By the time children are in reception, I suspect (but couldn't claim to know) that most will 'hear' the various sounds making up a word already. Reception teachers will be building on/extending that skill that many in their class will already have, at least partially. For many kids they won't actually need to 'teach' them to hear a particular sound; they will just have to tell them that this sound /s/ (which they have already learned to hear/detect in words; maybe a bit of reminding/reinforcing is needed) is represented by the grapheme . And I assume (again without knowing) that if a child in reception struggles to discriminate between sounds, e.g. dog and dot sound the same to them, or they struggle with identifying if the word 'cat' contains an /a/ sound, then the teacher would try to help them with this.
Be that as it may. I am not claiming that Montessori in general or this book in particular is 'right' and anyway may memory of what the book says may be poor. Just putting one approach out there.
I think that my own conclusion from my thinking on this thread would be that if I were to attempt to teach musical notation (which is only going to happen in a thought experiment :) ), I would start by finding out if my students had a grasp of that thing that the notation is meant to represent. If they did, all good - get on with teaching the notation. If not, I would try to find ways to help them gain at least a basic grasp, so that the notation then actually means something to them.
If you are thinking of musical notation as similar to numbers, i.e. the symbol carries the meaning, then perhaps you need to be researching how children learn to count/understand numbers/understand amounts, rather than how they learn to read! :)