Furthermore, the DMR-applied surface consistently showed a drag coefficient lower than that of the smooth surface up to the highest measured Reynolds number (3.6 x 10⁶)
This is the most surprising finding to me. The treated surfaces even exhibited lower drag in the turbulent flow regime.
Wait, is it also lower in the laminar region? Or is it higher in the laminar region, but laminar region lasts longer so overall improvement? Man a graph would be nice...