Can you use a helper column? It would contain a formula to identify which names were listed more than once, and they might be grouped as shown, and for all names where it was listed only once, they are grouped in a single "All Others" group.
It generally isn't good practice to show a sub-grouping that doesn't equal the original total. Often that is a sign of some sort of manipulation of data.
Could also just use a PivotTable. I prefer those over subtotals, as it doesn't mess up the alignment of my raw data, and I can easily sum/count/average any of the settings.