Needs more granularity, down to state, city, zip. If we had that level of granularity we might be able to establish a pattern of susceptibility based on carrier frequency.
It would be helpful to know if the problem is more prevalent on GSM-850, GSM-1900 (US/Canada), and/or GSM-900, GSM 1800 (elsewhere).
Distance from the cell tower would also be useful to know, because there is a db / signal "negotiation" between the tower and the cell phone w.r.t. how much signal boosting the phone should allow.
Collecting just zip codes or better yet, lat/long (e.g. from the Compass app), would at least allow AT&T to draw the proper correlation based on the tower data that they don't release to the general public.
I happen to be in Dearborn, MI. AT&T is on GSM-1900 in our area. The nearest tower is about 3-4 miles away and we suffer from "phantom" 5 bar signals that when actually using the phone will drop to 0 in the house and down to a barely usable 1-2, possibly 3 bars outside.
The behavior I am describing doesn't matter which AT&T phone I use -- it's more a measure of how bad the signal reception is. IN GENERAL, the poor signal HAS NOTHING to do with bridging the antennas on the iPhone 4, specifically. ALSO, even with good reception, e.g. 5 bars, loading of websites often comes to a halt on AT&T due to their "backend" feed to the internet often not being able to cope with oversubscription. The fact that the backend is not 100% consistent with page loading could be ANOTHER factor in the symptoms people are seeing.
I say all that just to establish a BASELINE for my own iPhone 4 experience.
In the BEST of cases when I have had a consistent 3-5 bars of signal, I have NOT been able to reproduce the issue with signal degradation due to pressing firmly with my palm, thumb, fingers -- whether moistened or dry, doesn't matter. No amount of repeated/consistent attempts has exposed any weakness, whatsoever in iPhone 4 signal degradation for my iPhone 4 or another iPhone 4 that I have tried from the same location under the same conditions. Whatever the number of bars I have at different locations from 3-5 bars, that level does not decline, no matter how I touch the antenna's -- also the page loading does not falter any more than TYPICAL with AT&T, whether or not I am touching the phone during the loading.
IN SHORT, it doesn't matter what I try, I cannot reproduce the issues that others have been describing.
So for my part, I can attest that on the GSM-1900 MHz frequencies, 3-4 miles from the nearest tower, at a time with the phone was consistently reporting 3 bars of signal strength, I am UNABLE to reproduce the issue with iPhone 4 signal degradation while bridging the external antenna's. This counts as TWO data points, since the same tests have been repeated consistently with no issues using two different iPhone 4's (same conditions, same location, etc.).
Perhaps GSM-850 is more susceptible?
Perhaps being closer to the tower with a stronger signal causes the phone to "negotiate" down to a lower signal boost, and in that mode the problem is more likely to surface? Yes, that's contrary to expectations, but could be the case, since being right under the tower will result in the phone lowering how much it boosts the signal, so MAYBE at minimum boost, nearest to a tower, the iPhone 4's antennas are more susceptible to the "bridging problem."
My main point here is that what frequency the phone is on and the distance from the tower, and numerous other factors are at play here. Which is not to say these would be the only factors. There may in fact ALSO be the possibility of manufacturing variances contributing to the problem.
