I found an interesting article (can’t speak to its accuracy) http://www.eetindia.co.in/ART_8800534325_1800001_TA_3d506bb0.HTM that claims some knowledge of the 3G iPhone. According to this article, the accelerometer is an LIS331DL from ST Micro (datasheet: http://www.st.com/stonline/products/literature/ds/13951.pdf) which is a 3-axis linear accelerometer. I note two things about this device:

1) it’s linear — but you’re reporting yaw and roll type data (correct me if I’m wrong) which would require more than one accelerometer to generate (or it’s making some major assumptions, possibly the accelerometer data is relative to the center of the iPhone?)

2) it operates in +/- 2 or +/- 8 g mode. My guess is the forces seen at launch easily saturate both modes and the iPhone most likely operates in +/- 2 g mode for better resolution for the most common uses of the iPhone.

I also dug up the datasheet on the GPS module, a Hamerhead II PMB 2525 from Infineon. (http://www.infineon.com/dgdl/PMB2525-Hammerhead+II-pb.pdf?folderId=db3a304316f66ee80117824fc0d71e07&fileId=db3a304316f66ee8011782518d4a1e08). This module doesn’t seem overly powerful and might explain the data lag.

If you’re interested, I probably have a GPS/cellular unit and some accelerometers with better range sitting in a box somewhere. If not, they can be bought for reasonable prices. I’d be happy to help build a custom electronics payload.