I think this isn't a bad idea at all, but comparing in the what-to-buy sense the value may not be that great unless it is found that different brands have very significant differences. This could be more useful for testing old battery sets to find out which ones are still good (for those of us who don't have a charger/discharger gadget to do that automatically using a constant load).
To properly simulate actual camera use, the script really needs to switch to play mode between a series of shots and other action and stay in play mode for most of the time (ideally the camera should be turned off and on but a script can't do the latter) or otherwise the batteries heat up more than they do in normal use and results will be unrealistic.
