As I recall, the original Safety and Contest certifications had certain performance parameters, like Safety was within 15% of stated thrust and impulse, and Contest was 5% within some amount of standard deviation.
Those went away sometime while I wasn't paying attention. Apparently, DuPont's black powder, which Estes depended on, suffered varied quality and efficiency in the 70's and 80's and Estes found it difficult to hold its motors to the same performance parameters as before.
So NAR certification simply meant they were consistent, worked properly and the company that made them could provide customer support (and a liability chain). Contest certification then became essentially the subset of motors that were commercially available to most competitors.
The liability chain is, apparently, the sticking point to making initial certification permanent.
Date codes on motor labels is probably another problem. If you say, we certify the Jones A7-5 motors made between 5-2003 and 8-2008, but no others, you've got to have the date stamped indelibly somewhere on the motor. But motor manufacturers have been inconsistent about that at best. Estes went to that 1970=A, etc system for years, then for awhile a code probably based on which Mabel it was made on, and marketing says some people think of the date as an expiry date. Aerotech had some octal system or some such, and for a long time they were printed on the ejection charge cap, so you couldn't tell anything about the motor after operation. And that's ignoring the whole reload situation.
So, what about doing away with certification totally? Well then, how do you count on anything the manufacturer tells you? You might trust Estes, but why? Why would you trust Aerotech? CTI? So you build or buy your own test stand and check them yourself, but that's usually impractical when you're talking about a K motor or larger, unless you've got money to... well, you know.