The '>20 W/mK' claimed by this method, while a lot better than the 5 W/mK of regular TIM, is still a fair way off the 80 W/mK that the article claimed for a soldered solution.