You bet it is. No matter how other people try to whitewash bitcoin, long confirmation time is the biggest problem of mass adoption of bitcoin right now.
I just went to a bitcoin ATM operated by robocoin. The other day I sold 1000 worth of bitcoins and I got 1 confirmation in 5 minutes.
the atm just needed 1 confirmation and it took 45 mins to get 1 confirmation today for 500 worth of bitcoins.
I think the problem here is more with a Bitcoin ATM which is designed to wait for 1 confirmation. There's still plenty of work to be done in the space of "fast payment verification" and I'm looking forward to see further market solutions (green addresses was one).
Deposit banks are one part of a solution. People would keep some of their bitcoins in a bank and the ATM would be able to accept transactions from the bank with no confirmations. This could be arranged to maintain anonymity.
A little more sophisticated: Payment insurance companies would allow people to pay with bitcoins quickly and directly. Such a company will charge a subscribing merchant a monthly fee and require the merchant to reject certain transactions it deems invalid in a short time window (say 5 seconds). It will pay out in full if any of the transactions it deems valid in that 5 second window are successfully double-spent. Consequently, the insurance company would have a strong incentive to make sure they are very good at detecting potential double-spends, while letting through as many innocent transactions as possible. They would be well-connected to the network and would probably spend a lot of money analysing the blockchain.
This isn't really a problem for the core devs. Besides, ensuring that sufficient decentralisation is maintained as the network scales is a more pressing concern.
Finally, while this is a problem for Bitcoin's adoption today, it is important to note that cryptocurrencies which have the same essential model as Bitcoin but simply set a lower confirmation time are not solving the problem, they are just adjusting the balance between this problem and that of orphans. I'm wary of this approach, simply because the orphans problem becomes more serious as the network scales.