Have some cake, you need it
I read quite a few gormless defences of Apple, but this takes the cake.
If your new phone drops calls *twice as often* as the old one, then that's a rise of 100%. Trying to debate that is identical to trying to debate whether 1+1=2.
Yes, I suspect that the rise is significant because I doubt that 50% of all calls are being dropped - that would be a failure rate sufficient to make the phone completely unusable. *Either* the rise in dropped calls is significant, *or* the iPhone was already an incredibly shit phone. You decide, you have to pick one of those, which do you prefer?
If you're in an area with good signal strength then the iPhone 4 is perfectly usable, no-one's suggesting otherwise. The problems arise in areas with marginal strength where the loss from antenna bridging lowers the signal below a functional threshold. This is also why data concerning the total number of dropped calls (a binary event) is useless as a means of comparing phones. What you want to know is the percentage rise in dropped calls in areas with marginal reception (which are still very common). If you make 198 calls in an area with good strength and they all get through, then make 2 calls in an area on the margins and both fail, then you only have an overall failure rate of 1 call per hundred. But that's scant comfort if you really needed to make a call in the second area.
Another useful statistic might be the average download speed actually achieved, since that's sensitive enough to track across a range of signal impairments.