I think this is a great question, still relevant now. This is an addendum to the answers already provided by Hobbes and iKrase.
Whilst I think the factors Hobbes has identified are spot on for the space industry in general the quality requirements on a cubesat structure basically come down to "don't get stuck in the launch pod" and "carry 1kg mass on very short moment arm from the primary structure".
We are discussing "price", rather than internal cost to the manufacturer of a cubesat structure. I suspect in today's market such vendors are is probably actually competing on price, and to an extent reputation, rather than performance. i.e. one might well expect it to be lower priced, somewhere.
Recall however that it is still in a market where the expectation is already that even cubesats should be expensive compared to other university/commercial equipment simply because space is expensive. Markets are not always efficient. A parallel is the hourly rates charged by commercial law practitioners compared to doctors, nurses, architects, engineers when the training and potential liabilities are comparable. People just accept, some less consciously than others, that this is the way the market is.
Judged by that metric, as iKrase has pointed out, cubesats still look cheap, just relative to much larger satellites that will operate for much longer.
Lastly, there has been a very loose trend1 that satellite costs often relate to less than being a factor of ten per kg away from launch costs. The idea behind it, whether it has come about by a logical reason or a coincidence, relates to launch risk and things being expensive because they are expensive. i.e. "you'd better spend more on getting the satellite (or launcher) right given how much is being spent on the other component". Obviously, it doesn't have to be that way.
1: personal observation, unreferenced!