I'm not sure that it's not worth talking about. Like Kev says, it's never going to be accurate cos of the amount of variables. But getting a bang-on figure wasn't what I was looking at; it was to do a poor-man's performance test and to clue you in to whether there, potentially, is a problem with a lack of power output.
If it took me 15 secs. to get to 60 on a straight, flat road using a stopwatch that started the moment I floored it and stopped at the moment I hit 60 then I (personally) would be thinking I've got an issue.
Just m.o. of course
Webby trying to measure the 0-60 for performance comparison with other cars may not worth the hassle.. but comparing to your cars previous performance may be useful .. but in any case the car will give lots of other hints.. starting with rough idle, black smoke, hesitation, stopping etc etc.. however, on the other side for small power losses I still carry the opinion that, on an auto it will be harder to understand and measure as the tranny trashes quite amount of power already..
although expensive, best way is to use the same dyno assuming they dont play with its parameters daily..