If you take the swarm concept and apply it to very cheaply produced pilotless aircraft, then the concept could be rather successful.
How many Typhoons do you think you need to hit on the ground after saturating local GBAD in order to make a significant impact? If you don't care much about losses, have sufficient weight of numbers that a decent percentage can afford to fail en route, or be destroyed en route, what then?
A suitable craft could be made for considerably less than $25K. Youtube is full of footage filmed by automated aircraft able to track moving targets on the ground, e.g. skateboarders, people on motorbikes. Applying the same techniques to attacking military targets cannot be far off.