Using the Battlog program, I ran benchmark tests on the Sprint 1.11.00F, USC 1.07.06F and Telus 1.04.05V. I do not yet have a complete set of data but I think I can conclude a few things. Signal strength varied between 95 and 97% and did not seem to correlate to a particular radio. I show only watt-hours because that takes into account battery voltage, assuming it was calculated correctly. Current draw can be misleading without knowing voltage but if someone would rather see the average current draw (or mAH) I will make a graph with those values instead.
One bug I noted in Battlog is that if you don't save the log right when it finishes, the phone data becomes invalid, as it continues to compute the values and includes them in the saved log. That is what happened to the second Telus test and why I didn't include the phone portion.
Are my values in the ballpark, particularly among people with "Excellent" battery life?