Boeing recommend adding half the steady headwind component to the Vref. The steady stream airflow usually begins about 2000 feet above the surface. Below that height, the wind speed is gradually retarded due to surface friction. It follows therefore that unless a speed adjustment is made by use of say the throttle, there will be a gradual bleed off with IAS once the aircraft passes from being in free-stream flow at 2000 ft.
Boeing recommendation includes the statement that head-wind additives should be bled off approaching touch-down such that the IAS will be at Vref at the threshold. By this statement a 15 knot HW component additive on top of the published Vref, would need to be gradually bled off from 2000 ft down. In practice, few operators actually do this. Some assume you will lose the 15knot additive in the flare. Not so - unless it is gusty, and then it becomes another story altogether.
If the autothrottle is used to control the IAS, then I understand that Boeing state that additives are not needed any more, because the autothrottle system will counteract any sudden bleed off and recover to correct speed. This assumes that the pilot cannot think as fast as an autothrottle system and needs additives to allow for his slower reaction to speed bleed.
I believe that one of the causes of over-runs on wet runway windy landings is that wind additives have not been bled back and the aircraft crosses the runway with excess speed. Surprisingly few operators bleed off the steady headwind component to cross the threshold at the correct height at Vref.
The above comments are centred purely around steady headwind components and do not address gusts.