"Also, people on these boards (especially those railing against grammars) hugely stereotype the intake of grammars. There's no doubt they're currently not representative, hence the drive for reform, but there are plenty of those who are not well off in the 'middle class' sense of the word."
No they don't!
Nobody here is saying 'all children at grammars are m/c, privately educated, non-white'. What they are saying is that some groups are very significantly 'over represented' and some groups are very significantly 'under represented', and that this view is supported by the evidence.
The issue which nobody in the pro-grammar lobby wants to acknowledge or address which is fundamental to this argument, is that children who form the cohort which makes up the grammar school intake also appear to do extremely well in non-selective schools, and that there isn't evidence that grammar schools make a big difference to educational outcomes for this group. This is what the Sutton Trust research says, and also what the research done by the Institute of Fiscal Studies says.
This is what the Sutton Trust has to say:
"Most of these analyses suggest that pupils in grammar schools do a little better than similar pupils in other schools, with the
difference somewhere between zero and three-quarters of a GCSE grade
per subject."
However, when they say 'similar pupils' what they mean is 'pupils with a level 5 or above in reading, writing or maths'. They don't control for sub levels or for the effect of parental motivation and input.
And even though they don't control for sub-levels, grammars still don't seem to make a big difference to outcomes!