Cambridge's efforts in this area illustrate my point about bias and poor judgement.
They set out to increase their intake of students from low SES backgrounds and from state schools. Over the course of a decade they increased the annual intake of POLAR4 Q1 students by a small number, around 50. They did much better with state educated students which was well publicised. However, from being on a par in terms of outcomes with privately educated students state educated students have lately performed significantly worse - privately educated students are 50% more likely to achieve firsts.
So, Cambridge - working on the false assumption that attainment is inflated for the well-off by schooling, rather than heredity and parenting - preferentially selected a demographic that appealed to their world view, and kept on preferentially selecting them year by year as they saw the attainment gap widening according to school type.
To a limited extent they got what they wanted in terms of diversity and inclusiveness but at the expense of outcomes. Does that ring any bells?
edit - annual intake