Looking back at it now, our argument for leaving the rails via-less to allow for flexibility no longer seems as strong. I'm going to discuss with the OSU group, and if there's no objections, I'll add vias inside the cells, spaced 0.17 / 2 = 0.085 um away from the outside edge of the rails, or in other words 0.305 - (0.17 + 0.17/2) = 0.05um away from the inside edge of the rails. That should maximize the number of vias we can paint in before the cells are abutted.