he can remember the sounds he has decoded in a single word, but when he has to decode 2 consecutive words ( in reality it would be more like 5) you have to remember the previous one whilst saying it, at the same time as decoding the next - and I think this is where the problem lies - is it called working memory?
I don't quite understand why he has to remember the word he has just decoded. Once he has said it he can move on to the next word. He will be quite halting at first because he is learning, but, as I said, the more often he decodes a word the quicker it will get into sight memory (this is a totally different process from 'learning sight words) and he will improve in fluency.
perhaps there isnt a solution but I was hoping some of the techniques used in a UK classroom would have been transferrable
There is no magic key to teaching dyslexic children. They may need much more practice to achieve automaticity, but the body of knowledge to learn is the same, whatever their problems. If your son had a problem with discriminating the sounds in words you would have to focus more on training him to 'hear' them, but I understand you to have said that his phonic knowledge in English is OK, so such a focus doesn't appear to be needed.
I find that most struggling readers do have good 'phonemic awareness' (i.e they can discriminate the individual sounds in words) but the most 'challenged' children have difficulties with rapidly and automatically 'naming' the sound of a letter, or letters. This obviously affects their ability to decode rapidly and the only way round this is to practice, practice, practice the correspondences until they are absolutely automatic. Once they can do this they are able to sound out and blend words easily; the next problem comes with those who take hundreds of repetitions to get the word into 'sight' memory. This may take ages, but in the meantime, as I said earlier, so long as they can sound out and blend rapidly they will be 'reading' the word (however laborious it may sound to an observer) and, so long as the word is in their spoken vocabulary they will understand what they are reading. Obviously this last point will not apply to your son, who doesn't understand Arabic.
I do sometimes encounter children who can decode and blend and do know their correspondences but for some reason look at a letter and say something entirely different from the sound that letter represents. I don't know if this is because they have not previously been rigorously taught to 'say exactly what they see' (as I don't work with them until Y7!) or because there is some very real processing problem involved. It is not a problem which I have ever seen discussed in the literature on dyslexia and it is one which defeats me. Apart from insisting that that they say exactly what they see I can't find a way round it!
There is a mystique surrounding 'specialist dyslexia teaching' which quite unfounded, but does impress people! The Orton- Gillingham system, on which most 'dyslexia programmes' are based was developed in, I think, the 1930s and it basically nothing more that systematic phonics teaching, albeit taken extremely slowly and in an unduly complex way. It features 'multi-sensory' teaching, which has been elaborated to an absurd degree from the multi-sensory 'seeing, saying, hearing & writing' procedures advocated by the early developers of O-G. There is no research evidence to back the wilder flights of fancy, such as feeling letters in a bag, or having them written in a finger on the pupil's back, or making them out of clay etc. etc. Advocates of MST know this. If I had time I could link you to an MST advocates website which has a very recent appeal for research into MST.
What I am trying to say in this rather lengthy fashion is that there are no magic techniques used in English classrooms over and above learning the correspondences to automaticity and how to sound out and blend the written word to produce the spoken word.