I think you read it wrong. The table showed the number of marathon wins by British women and the number of those wins that were sub 2:48 not total runs (winning or not) that were sub 2:48.
however, a slight amendment to your explanation gives us:
The table shows the number of podium finishes in major marathons and the number of other wins that were sub 2:48... (which explains why he adds them together in the final column).
and that now makes some kind of sense. but the table itself is still wrong.
Dani Nimmock won Manchester in 2:38; Helen Davies won Brighton in 2:38; Natasha Cockram won Newport in 2:44; which I make three wins under 2:48.