Back then, some brands would rate their equipment at what ever specs would give the best result, so their stuff would look better than the other guy's. Out of a given number of ratings, some might be rated at 4 ohms, some at 8, some at just a single frequency or a limited range, some at RMS watts some music power, some at high distortion levels. On top of that, a given maker might rate distortion at 1 watt, power at 1khz, and frequency response without regard to either of those. For the most part, the figures were truthful, but they could be manipulated to give the impression that their performance was better than it really was. The result was that as a consumer, it was difficult to compare specs between makes because they weren't based on a standard measurement.
What the new rule did was require each manufacturer to include the specs measured across a given frequency range into a standard load with both channels driven. As I recall there is also a set time period that these specs must be maintained. They were still free to publish their other measurements as long as the standard was included. There was some merit to some of these ratings such as 'music power' which gave some idea of the amount of reserve power that was available for peak transients.
After the 'Power Wars' died down, most makes seemed to be content to just publish the standard ratings. With the advent of BPC, it does appear that they are tending to advertise the non-standard specs more and more. AFAIK they are still required to provide the RMS rating, but it may take some effort to locate buried in fine print.