"why are the countries which consider all children can learn, and expect all children to make progress the ones who do best in international tests? China/Japan/Finland."
They don't believe 95% of their children can reach some high level and they don't. Their PISA bell curves are just a bit further to the right and in some cases stretch out much further to the right which suggests their top children have fewer problems with ceilings.
Finland doesn't look so good these days and you can find Finn's talking about problems with their top-of-the-range children. TIMSS (maths) says Finland is essentially at the same level as England.
China is actually Shanghai in these international tests i.e. essentially testing higher IQ children and claiming they represent the entire country. The average Singapore IQ is significantly higher than here. The Japanese average IQ is also higher. Ditto for Korea. There might be a clue here, even before we get to maths, language and working memory.
We know English academic achievement is strongly influenced by a roughly equal share of genes for intelligence and genes for behavioral traits. Intelligence and hard-work. Different culture, some of which I expect is rooted in genes for behavioural traits, obviously applies to all of those countries e.g. children working a lot harder in their own time, not being quite so 'fighty' etc.
It was Hannah's Sweets most recently. Please do read the Thompson link I posted earlier because they know a lot more about cognitive science etc. than we do.
--
"Teaching other children while the teacher teaches less able are norm for my ds"
12 year-old DD's maths top set is still mixed ability and she sometimes gets asked to go help explain things to the weakest children. What is probably the national top ~20% is still quite a significant spread of ability and it sounds like that peer-tutoring rarely works. She apparently can successfully explain things to a child much like her though.
--
"as most of the damage it does tends to impact on children who aren't on the top tables"
Ah. The inevitable built-in bias in the polarity.
Can you point to any credible, independently replicated evidence showing a beneficial effect in having lower-ability children in a mixed-ability class in a typical English classroom?
Your 'damage' is definitely not proven, not least because there is no credible domestic research and for the most part this claims are founded on 27th-hand takes on some quite dodgy stuff based on other dodgy stuff and imported from the US as usual. The compelling theory for this is that relevant research doesn't get funded here because it risks undermining some of the cherished ideology common to the majority of the purse-string holders.
But the EEF have finally funded just a little something in this area so we might get to see. Perhaps it has changed since I looked, but the EEF also had a page on mixed-ability which (presumably because someone got hissy) mentioned 12 months progress from some G&T programs i.e. that might make a rough measure of how much of their potential is thrown away for the greater good every academic year.