That's not actually what this paper says. Reduced neutralization in a lab experiment doesn't necessarily imply zero protection in the real world. (As we're now finding out.)