On the New America Foundation data:
This source has gotten a great deal of mileage recently, including articles in the New York Times.
But if the actual goal is comparing jihadist-inspired murders in the U.S. with right-wing-inspired murders, then isolating the data on ideology to "domestic" or "homegrown" offenders really has nothing to do with it.
But if you don't do that tweak, then 9/11 would count, and the whole thesis collapses, even if you run the endpoint back to 1995 and Oklahoma City.
I would make this point on New America's site, but their "feedback" link doesn't work.