BUT..BUT..BUT...

"That can't happen because it's not in our model."  Brad DeLong observes that the facts Piketty accounts for are assumed not to exist in American neoclassical growth-economics:

One thing that I had not fully realized before yesterday was just how much heavy a lift Piketty has in trying to persuade the American neoclassical growth-economics community within economics departments. Their--our--default view of the world is--very strongly--that it is characterized by a Cobb-Douglas aggregate production function, in which the rate of profit moves inversely with the L ratio and in which as a result the capital income share of total income is constant.
You may say: but if you have a model in which you assume the capital income share is constant, you then have no chance of ever explaining fluctuations in income distribution. How can you use a model in which fluctuations in income distribution do not happen to criticize anyone trying to explain why they do? And this is a more than fair cop. But that the habit of thought is not rational doesn't keep American neoclassical growth-economists in economics departments from doing it: their--our--first reaction to Piketty is:  That can't be right, because in our model the capital-income share of total income is invariant to shifts in the wealth-income ratio." And they--we--typically do not take the second step in the argument and say: "wait a minute: in our model nothing causes shifts in income distribution, so we need to model"...

DeLong also has good links to other remarks by and about Piketty.