When you're arguing policy, you have to decide whether you're arguing in the perfect world or the real world, and then you have to hold that constant. There's no use arguing a textbook policy against a piece of legislation, or vice-versa.
You see this pretty clearly in debates between supporters of cap-and-trade and supporters of a carbon tax. If you're interested in determining the elegance of the policy, you can compare a textbook cap-and-trade program to a textbook carbon tax. If you're interested in determining the likely real-world shape of a policy, you can try and game out the likely compromises and horsetrades that would allow cap-and-trade legislation and carbon tax legislation to pass through through Congress. But what you can't do is compare the cap-and-trade legislation to carbon tax theory. These things are not alike. But that's the argument you tend to get.
Which gets at the problem with Andrew Sullivan arguing that a carbon tax is superior to cap-and-trade because it "will bring green benefits without the kind of crude regulatory scheme that could stigmatize environmentalism for a long time." As Kevin Drum explains, there's nothing crude about cap-and-trade in theory. It only becomes crude when shoved through the legislative process. But anyone who thinks that a carbon tax would navigate Congress with any less damage should look at the income tax code. Like the carbon tax, the income tax is clean and straightforward in theory. But no one examining it today could deny that it has grown crude and complex.