I think the objection to statistics about immigration (or minority populations) and crime is that they are inherently misleading because they generally compare crime rate to the population as a whole (as you did, jumping to the per capita), rather than to equivalent socio-economic groups. Consider two headlines: "immigrants commit most of the violent crime in this country" vs "most violent crime in this country is committed by poor people, and most immigrants are poor". How you phrase the headline makes a big difference in how people might think about policy.