I agree with the basic claim of this response: economic inequality itself it not a problem. I don't really think I would care too much about an extravagant billionare class if the lower classes had 100K household incomes in a robust democracy where the proletariat would have have fair representation. The problem of poverty, generalized, is a problem of access to essential resources. What counts as "essential" of course might be debatable, but the metric for improvement should be about reducing the fraction of the population that doesn't have access to essential resources. However...

We will not do a good job of fixing the bad causes of economic inequality unless we attack them directly.

I would replace "will not" with "may not". If the "good" and "bad" forces at work are sufficiently correlated in their ultimate causes then a blunt metric of reducing overall inequality might turn out to be a nice strategy.

EDIT: In particular if per capita income is known then the total variance and behavior in the tail may place strong constraints on the fraction of the population below a threshold.