Are stupidity fees economically efficient?

Ryanair's CEO was recently quoted as saying that customers should have to pay exorbitant fees for being stupid, after a woman complained about paying roughly $380 to have boarding passes printed for her family. The discount airline charges 60 euros to print a boarding pass at the airport.  

What is a stupidity fee? Charging 60 euros to print a boarding pass is one example; extremely punitive charges on late credit card payments might be another. I would define "stupidity fees" as having the following characteristics:

• Stupidity fees are charged to a subset of people who purchase a good or service. They are avoidable if the consumer takes certain actions, but once the consumer is at the point where the fee is actually applied, it is very difficult to avoid or re-negotiate.  For example, once you are at the airport with no boarding pass, it is difficult to avoid paying to have your pass printed. 
• Stupidity fees are higher than the cost to the supplier that results from the consumer's failure to take the action. For example, the total cost of printing a boarding pass is far less than 60 euros. 

The result of these characteristics is that stupidity fees often serve to cross-subsidize those who are savvy enough not to pay them. This allows the supplier to attract customers with lower prices or perks that they could not otherwise afford to offer. 

So are they efficient? Cross-subsidization is not necessarily inefficient.  However, I see three potential problems with stupidity fees:

1. People tend to overestimate their own intelligence and savvy, and therefore probably don't think they will fall prey to a stupidity fee.  by underestimating the likelihood that they will pay a fee, buyers underestimate the cost of stupidity fees, leading them to accept deals they might not if they valued the cost more accurately. (This is an asymmetric information market failure.)

2. Because the fee far exceeds the cost to the supplier of the stupidity, the supplier has every incentive to encourage the stupid behavior. As a result the supplier doesn't have a good incentive to give consumers clear information about how to avoid fees, or create systems that help consumers avoid them.

3. The cross-subsidization may or may not be progressive. I see arguments for this to go either way. On one hand, those with more money may be more willing to pay fees in return for convenience, whereas those with less money may be more careful to avoid fees. If this is true, the cross-subsidization would be progressive. However, if the fee truly is a "stupidity fee", people who are less educated, or purchase the service less, end up cross-subsidizing more