It's important to note that correlation doesn't imply causation.
Also, greater accidents doesn't imply greater design flaws. The data here is only concerning accidental damage. However, it doesn't and cannot isolate the causes of accidental damage. In other words, we can't tell from this data if the design is contributing to increased incidence of accidental damage, or if greater carelessness is contributing, or if it's just statistical creep in the fact that as the iPhone audience becomes even broader, people who previously never owned a smartphone and didn't think to take care of one are experiencing accidental damage at a higher frequency due to greater likelihood of carelessness.
These are all separate metrics that need to be isolated out, and weighed against actual field tests assessing the likelihood of damage in identical conditions.
To wit: I have damaged four iPhones prior to the iPhone 4. I have not damaged the iPhone 4 once. After damaging four phones, I'm much more careful with this one. Even so, I've definitely dropped it a few times and not seen the screen shatter yet.