It's becoming well known that the 2.4 Ghz radios may be susceptable to undervoltage issues on the receiver battery supply. And, these have resulted in loss of control of the model. Spektrum did do some software updates a while back to reduce the time of "Rebooting" in case of a low battery issue. My testing on updated AR7000 receivers shows this "rebooting time" is less than a second.
IMHO, this is not a receiver issue, it's a problem with the receiver battery selected. By actual test, I've measured the peak currents pulled by all 7 Hitec 645MG in my giant scale Extra 330 model. That peak current was measured at 14 amperes, with the one millisecond peak holding feature of my Fluke 87V digital multimeter.
So, to illustrate this issue, two discharge current tests of 12 amperes were applied to a new LiFe 3200 Mah battery, and to a new 5 cell 2700 Mah Nih "AA" type receiver battery.
The attached graph, taken from my Western Mountain CBAIII battery test unit shows it all.
Take a look at the attached JPG: The 12 Amp load on the Nih battery was shut off when the voltage dropped to 3.6 VDC, a value reached in about 2 seconds. Note that those Spektrum receivers will reboot when the receiver battery voltage hits 3.2 VDC. The Nih battery test was conducted immediately after it was fully charged. The results will be worse after a few flights on the model.
Again, IMHO, the 2.4 Ghz receiver of any large model with more than 4 servos and a receiver battery should be using a higher powered receiver battery than those "AA" type units. Recommendations would be one of those LiFe types, a "Sub C" 5 cell Nih battery, or a uBEC such as Castle Creations 10 (or 20) Amp Switching Battery Elimination Circuit.