I think 'to do' differentiates itself from 'to be' in E' because while people use 'to be' to make global declarations of property (e.g. "That movie is bad" vs "I didn't like that movie"), they use 'to do' to make global declarations of action (e.g. "The IRA does bad" vs "The actions of the IRA reflect violence").

While avoiding 'to do' additionally aides in specificity, I believe avoiding 'to be' has a bigger impact, and avoiding both near impossible.