First, here's the abstract(which is all I can read with out a subscription):
ABSTRACT Most of the 3-mm-thick globally distributed Chicxulub ejecta layer found at the Cretaceous-Tertiary (K-T) boundary was deposited as condensation droplets from the impact vapor plume. A small fraction of this layer (<1%) is clastic debris. Theoretical calculations, coupled with observations of the coarse dust fraction, indicate that very little (<1014 g) was submicrometer-size dust. The global mass and grain-size distribution of the clastic debris indicate that stratospheric winds spread the debris from North America, over the Pacific Ocean, to Europe, and little debris reached high southern latitudes. These findings indicate that the original K-T impact extinction hypothesis?the shutdown of photosynthesis by submicrometer-size dust?is not valid, because it requires more than two orders of magnitude more fine dust than is estimated here. Furthermore, estimates of future impact hazards, which rely upon inaccurate impact-dust loadings, are greatly overstated.
It's clear from the abstract that he's creating a distribution of grain sizes. The pure fact of physics is that different sized dust grains scatter and reflect certain wavelength light more effectively than others. In this case, he's finding more coarse grains than fine grains, and he's finding 100 times less of the fine grains necessary to reflect the visible light that plants use best to grow. Pretty straightforward stuff, and more complex then earlier models.