If there's one thing libertarians love, it's redistributing money through the government.
For things like transportation I'd much rather see a "miles traveled" tax than a carbon tax (it could even be tiered for electric vehicles). It's far more specific, and thus harder to lead to crazy far reaching taxes in tangential areas.
Why? The harm done is relative to emissions, which is exactly proportional to the volume of fuel going in. A tax on miles traveled would unfairly punish cars with great (or infinite) gas mileage. (Unless you think the relevant externality is cars on the road...)
But they're dwarfed by non-environmental factors like road wear-and-tear, and accidents, which are covered by other taxes and insurance.
The past 40 years of politics shows that this is simply not true in reality, though it is often repeated. There is significant political will to lower taxes, and taxes get lowered frequently. More frequently on the wealthy, but frequently nonetheless.
When it comes to taxing externalities and improving market function, there's often significant room to improve the market. As long as costs can be estimated within ~50% it's going to be a big win to legislate them. However, since the cost of going over is often not linear, it can be difficult to enforce in a fair way in the market without some sort of auction scheme like what goes on in cap & trade.
A miles tax is far worse than a carbon tax; it's far more of a stab in the dark. It doesn't even account for wear & tear in any way, which is dependent on an polynomial of the vehicle weight.
Large fossil fuel companies, like Exxon, are planning on an $80/ton carbon tax. There's zero reason to not phase this in along with the Paris accords. It's silly to ignore the work that those, negatively affected by the tax, have already agreed is going to happen.
Source: http://www.justfacts.com/taxes.asp (tons of footnotes there to bea.gov
I would assume that the tax rate would be based on vehicle class, but that seems too easy.
Some example reading to open your mind: https://mises.org/library/externalities-argument
Argument that "it's hard" is not much of an argument at all. It does make me think that there's considerable flexibility in a "libertarian" position on externalities though.
Fine it's impossible to correctly calculate externalities. I don't see the line from there to: "BigCo dumped a ton of waste in my pond, but it's unfair for the government coerce BigCo to compensate me because there's a chance the I might get more than the true value of that pond."
But I do agree that the best pricing schemes involve settling on price through transaction. Take, for example, cap & trade schemes. This is a widely used mechanism to manage negative externalities.
Externalities exist, whether or not the "Austrian" school has a way to deal with them. Considering the issue, deciding that one's philosophy can't deal with them, and therefore deciding to stick one's head in the sand instead is why mises.org remains fringe. It's an irrational, religious framework.
Read Mises Human Action and you will understand.. Austrian economics is a priori (built from first principles) not a posteriori (empirically derived)