Per the measurements made by Garfield, you aren't really gaining anything by switching the camera off, rather than just letting it idle. In fact -- particularly if you have sensor cleaning set to engage each time you switch the camera on and off -- you may be discharging your battery more than you would by just letting the idle timer do its job after each 1 minute period of no activity.
I disabled auto sensor cleaning for that very reason. When I was in Nepal for three weeks last year without ready access to power for recharging, I found that doing this and shutting the 5D3 off after shooting seemed better than letting it auto power-down. Then again, perhaps shutting it off just prevented me from tinkering with it and needlessly engaging AF. The CIPA standard (found at http://www.cipa.jp/english/hyoujunka/kikaku/pdf/DC-002_e.pdf
) assumes one shot taken every 30 seconds, with a powerdown/powerup every 10 shots. The 5D3 is CIPA-rated at 950 shots, so that equates to approximately eight hours of continuous use. I was able to shoot a few hundred shots over a few days on one battery, so it doesn't seem so bad in comparison. Interestingly, the CIPA standard doesn't prescribe a power off time, they just say it should be an amount of time which doesn't change measurement results. So to the OP, it looks like the camera companies may measure off-mode battery drain, but only to determine at what point it isn't detrimental to the battery performance that they report.