Antenna extension

The optimal length of an antenna is determined by the wavelength of the radiated frequency. Cell phones use higher frequency bands so the antenna length requirement is for a short antenna. If you increase frequency, you decrease wavelength.

An example is CB radios are lower frequency bands compared to FM radio and therefore have longer antennas. FM radio antennas have an optimal length of 31 inches based on their freq spectrum.

So the best approach for the antenna is use the one designed for that specific application. I would not expect a casemaker to be better at this than the phone designers. (queue the wisecracks)

Exactly.

From an RF engineering standpoint, part of the issue with the external antenna is that the user's skin (which is conductive) bridges the 850/1900Mhz GSM cellular antenna with the 2.4Ghz WiFi/BT antenna, changing the length to the sum of both antennas and thus changing the RF properties of the antenna, effectively "detuning" them and causing the signal quality to deteriorate. At least in theory, this conspires with bugs in the new signal acquisition logic such that is isn't able to re-lock onto a signal causing the now infamous drop from "wonderful" to "pile of poo" to "I see nothing" on the signal strength meter...

IF (and it's a huge "if") the case's embedded antenna was properly tuned (sized) for the GSM frequencies then sure, it might help, but
A) it would be a non-issue since the user wouldn't be bridging the antennas in the first place since the case would be providing insulation
just as you stated
B) I wouldn't really trust "random plastic molding company" to actually get the RF design right anyway.
and finally
C) Since the case would be an "antenna" it would be subject to all sorts of regulatory issues and testing by the FCC and other International communications regulatory bodies which would increase the time to market and the costs of both engineering the device as well as impacting the final retail price (it wouldn't be inexpensive).

RF is a black magic to begin with and the RF engineers the mages, it's not a simple matter to design a proper antenna. Do it wrong and you make signal quality worse as we are seeing when we hold our iPhone 4's "wrong".

I don't think you need an antenna extension in the case. All the case needs to do is to prevent the current problem.

The current problem is a bridging issue. On the lower left corner of the phone, that is causing the problem, there is a seam where the two pieces of metal meet. If you hold you finger on one piece or the other there are no problems. It's only if you bridge the two pieces that the problem occurs.

You can test this yourself. Hold your finger above the black seam line and you won't see any bars drop. Do the same below the black line and you get the same result. Hold your finger so it crosses the black line and you will see bars start to drop in a second or two.

All a case (like the 'bumper') has to do is prevent the user from bridging the two pieces and there will be no problem.

Just about any case that covers the side should prevent this.

Amen! This makes sense to me. Regarding the frequencies, is there a way to "tune" the firmware to look for a broader spectrum? Is it just tied to the length of the antenna?

Has anyone tried the Full Body Covering from ZAGG for the iPhone 4? I've written to them 3 times but they must be really busy, no response. "IF" the problem is connectivity between the two antenna would this "skin" actually keep the contact in check? Of course the iPhone would have to be left off for 2 days (48 hours) while the covering cured, but once it was in place I'd think it would do the trick. Am I off base here?