All of those things below. Actually as hobnob has pointed out the research methodology of the sutton trust report quoted
"The findings indicate that schools with similar average A-level performance vary considerably in the proportions of students admitted to the Sutton Trust 13, and Oxbridge, universities. Crucially, some schools ‘over-perform’ in terms of their hit rates given their average A-level results, while others ‘under-perform’ – sending fewer pupils to elite universities than would be expected given their average academic results."
suffers from an epidemiological ecological fallacy as individual achievements and outcomes are not compared, only outcomes at school level with averaged out A level scores.
The data in that report are also outdated as they precede the A star grade and also the use of school weighting systems by Oxford and Cambridge. Updated research using logistic regression would be helpful
If one reads through the other Sutton Trust reports , one is presented an overall picture of 'top' (usually independent schools) feeding into 'top' universities and then producing 'top' movers and shakers in all areas of society.
Not a problem except that it would appear that many bright children not going to independent schools might not then go to top universities and not get the chance to be movers and shakers.
Which means that, if the trend continued, those at the top would continue to be socially remote from the average person.
So that is why the change I would like, if this trend is shown to be persistent, is to try and ensure equal opps at the university admission stage. A point with which I am sure you agree.
I think you huw and hobnob are saying that equal opps is happening because a child with 3 As has the same chance of getting in regardless of school attended.
I am not convinced this has been robustly demonstrated.