I don't think it is fair to dismiss their discoveries as 'non-issues', but rather they should be thought of as techniques limited to criminals with significant resources. However, nearly every technological problem becomes easier with time, and these attacks should be addressed before they become a major source of fraud.
I agree with this. I could think of nothing more frustrating than having the issuer deny me a chargeback because of the assumption that there is no card present fraud with chip-and-PIN. The possibility is remote, but it is nonzero. I still maintain that public perception in the US would keep zero liability policies in place, but I think there will be additional scrutiny of fraudulent card present transactions in the future.
Additional scrutiny might not be a bad thing because it reduces a moral hazard that now exists. Right now everybody in the US seems to take a cavalier attitude when it comes to credit card security. Consumers know that they're protected from any fraudulent transactions, so they make little effort. If a card gets lost or stolen there isn't the immediate need to cancel it. Merchants in the US rarely check for signature (or ID - even though this violates Visa/MC policy as long as the card has been signed), and why should they at this point? Someone can just clone a magstripe card, so it's impossible to tell if it's even a legit card. Issuers pretty readily courtesy credit small transactions and will remove any fraudulent ones. Everybody seems to accept a certain amount of fraud in the system as it currently stands without wanting to do much about it.