Yeah, but even low carbon steel has about a 3,500kg/cm^3 shear strength.
The Shimano spindle failures I've seen invariably occur around the non-drive spindle arm spline interface. There's no doubt it creates a weaker point on the spindle, but whether or not that's a tangible weak point compared to the other bonds involved is hard to say. They do caution on the "drillium" and do note that the readings will go haywire on the power meter if there's a fatigue fracture forming, so the main concern is a catastrophic failure...and I'd bet it will fail on the non-drive arm well before it fails on a 2mm hole should such forces be achieved.
Still, I think the potential market is pretty small, but it piqued my interest.