The ones wanting to abolish he grammar school system are trying to drag those bright kids down to the failed the grammar level.
So in successful comprehensive systems, are the brightest children being failed, and do the most able pupils in such comprehensives only achieve what children in secondary moderns in e.g. Kent do?
If that were to be the case, then selective counties would enormously outperform similar but comprehensive counties - because once you match all the secondary moderns with the comprehensives who supposedly are only allowing bright children to reach a secondary modern level, then the pupils at all the grammars in the selective county would be outperforming the lot, bringing the selective county out miles ahead.
Except no, that's not what the data says. Within the limitations of matching any areas - variations in demographics, employment patterns etc - the data for selective counties and comparable comprehensive counties is the same. So to a first approximation, if you take 10,000 children of matched socio-economic and family backgrounds in selective county A, and match them with comprehensive county B, put them through the schooling system in each county (careful division at 11, or not) and then look at the results, they are statistically very similar. There is no overall benefit, at that cohort level, of having a selective system