Two things come to mind:
1. They may have managed to perfect the debayering technique so that there are no artifacts.
2. If the
first tests of the removal of microlenses and CFA can be believed, they actually show 32% increase in QE for red and 12% and 6% decrease for blue and green respectively (when compared to the colour version of the same sensor). If true this is actually quite good.
At this stage I would like to see more independent tests and real images before making any conclusions.