Originally, the expectation was to continue using chemically amplified polymeric resists. However, their roughness is quite severe. Metal oxide resists or metal-containing resists have been proposed and evaluated with their greater absorption. Now, Irresistible Materials also proposes fullerene-based resists at SPIE. It seems all these newer resists are going to be used at >20 mJ/cm2, so ASML's throughput projections will have to be discounted accordingly. With 210W, they must be using a noisy 17 mJ/cm2.
Their throughput (WPH) claims are tied to the 20 mJ/cm2 assumption, but on the other hand, this depends on the resist used. Without pinning the resist down (and there are many choices), ASML can't really make that claim.