It'd be helpful to see this study normalized for ROI over time, which is arguably a better measure of VC success than exit size.

Doing so might very well make Fred's contention hold up.