Over the weekend, I took some measurements, broken up by trip. I didn't include any of the time the car was parked (though I haven't noticed any large losses while parked.) One difference was that I did a lot of freeway driving this weekend, whereas my weekday measurements were for 100% surface street driving. The other difference was that the weather was warmer (around 90 degrees), though it was never cold for any of my measurements. (This is San Diego!)
The discrepancy between rated miles and energy was only 0-10%, and for the longest trip, 29 miles, mostly freeway, I used 40 rated miles and the trip meter stated energy use of 11.4 kWh, for a discrepancy of only 1%.
All the numbers are in the wiki in the second post of this thread.
I'm not sure what this means. My driving was probably more efficient, but that shouldn't affect the discrepancy between rated miles and energy. Perhaps even at 60-70 degrees, there's a lot of energy going to warming the battery pack?