There are probably dozens of parameters you could use to order political groups, leading to an unmanageable space for visualization purposes. But not all the factors are equally important, and I agree with you that degree of violence is perhaps the single most important organizing factor. It's a bit of a duh, since we want to prioritize for the worst potential consequences.

Course, what would happen if such a calibrated list revealed that a certain ideological leaning correlated with a propensity for violence? Not that I'm naming names (and anyway, everyone knows which leaning has the firearm fetish), but it could be controversial...and I wonder if DHS put out a disorganized list to avoid the political firestorm that could accompany the analysis implicit in an organized taxonomy of domestic extremism? This "let's not dwell on the past" retreat from a political fight has been my big disappointment with the Big O so far.