I don't see what the problem is. The information is comprised of basic GPS, microphone audio, and phone radio data that's very obviously being collected purely for debugging and diagnostic reasons. From the article*:
The phone collects the information, storing it in a file called "/media/.NSAquiredData" until it can be transmitted to the Motorola server at bmailvctms.gomoto.com, and comprises of the following:
* Number of dropped calls in the last 24 hours.
* Location data, sampled at 5am, 11am, and 6pm
* Location type of above (eg residential, business)
* If the location at 5am != 6pm, and 5am and 6pm are both residential locations, and 11am is a business, then:
- Whether 6pm is associated with a phone number that is frequently called but not marked "HOME", "FAMILY", or "WIFE"
- Whether a random, five minute, audio sample taken between 6pm and 6.30pm matches patterns marked "KISS", "WORD_LOVE", or "WHIP"
- Whether that audio sample contains both male and female voices, and whether, upon analysing a similar sample taken at 9.30pm, one voice matches but another voice does not.
* The date and time and location of any dropped calls
* The temperature of the phone at the time the calls were dropped
* The status of the humidity sensors at the time of any dropped calls
Seems perfectly reasonable to me.
* No, not that article, the other one.