Did Mankiw miss the point of the double dividend debate?

Stil another guest post this week from Jim Roumasset:

I don't see anything wrong either Mankiw's 8/16 NYT piece, wherein he discuss the merits of various carbon policies, nor his follow-up Mankiw's Economic View piece, as far as they go. In particular, he has not missed the point of the double dividend debate. Environmental taxation was once thought to potentially produce two dividends -- the first from reducing the deadweight loss associated with excessive pollution and the second from the reduction in labor taxes afforded by recycling the Pigouvial tax revenue. General equilibrium economists pointed out that this view neglects a third negative effect. Increased costs of the "dirty good" may exacerbate the excess burden of the labor tax (the "tax interaction effect") by more than the second dividend. It was therefore asserted that the optimal environmental tax could be less than the marginal damage cost, not greater as often assumed (see also the references in the comment section of Brad DeLong's Note to Self: Greg Mankiw's Anti-Environmental Bill Argument Once Again).

Mankiw's advocacy of carbon taxes over permits was in the propensity of the latter's being given away. His point is that more tax revenue from the full carbon tax is better than the smaller revenue from a permit system with "giveaways" because the additional revenue can be used to reduce the payroll tax.  The tax interaction effect will occur with either taxes or freely allocated permits. (It's not the revenue that causes the "tax interaction effect" but the increase in the producer's price of labor and the consumer's price of the dirty good.

To summarize the scorecard to date:

1. Carbon taxes are better because of the latter's propensity to be given away, foreclosing the  opportunity for revenue recycling.
2. Carbon taxes are better for dealing with one-sided uncertainty about the marginal benefits of  emissions because the marginal benefits are steeper than the marginal damage costs.
3. The primary advantage of a permit system is that it lets the political system decide how much  to give away.  (We could do the same with block-rate taxes, but that would never get off the  ground.) Unfortunately, the political system doesn't stop with what percent of their efficient  emissions all polluters get. Everyone gets something different, including consumer groups, as a  result of rent-seeking, and the lobbying efforts dissipate at least part of the potential rents.
4. Neither taxes nor permits can beat hybrid plans, once uncertainty is introduced. But the  literature is incomplete regarding the optimal hybrid under two-sided uncertainty.