I've heard enough "cannot reproduce" reports to suspect it's a minority of devices, still that's a lot of people affected. My suspicion is that there is a coating process that is failing in some cases, which would mean it could probably be resolved for future production runs.

I've only been able to reproduce it in areas where the signal was awful to begin with. In areas where my 3gs drops calls weekly, using other people's 4's I can reproduce a dropped call with some regularity. My guess is that it's so hard to reproduce because even by attenuating the signal by holding it wrong, a lot of users are in an area where the signal is strong enough that it doesn't matter.

Unlikely, according to this (posted to HN eariler today)

http://mashable.com/2010/06/29/leaked-iphone-4-docs/

They seem reluctant to give cases away, which is a bizarre move, imho. A few thousand or tens of thousands of cases would sweep this ongoing PR bleed under the rug, simple-dimple. That is, until the lawyers get into it. Then it gets legs of its own, can go on for eons, and there is the very real possibility of a document subpoena revealing some mention of the possibility of the problem by an engineer which can be spun up into punitive fines.