Most service weapons get locked into their technical data packages and can be arduous to reform any flaws once adopted, sure. The 92, however, has been exceptionally fleshed out of its shortcomings through its service lifespan. The Army has fired over 3 million rounds through them studying their reliability, without the aftermarket changes. They expected for the last M9 a service life of about 25k rounds, but the data dictates roughly 18k is the real life number. That was before the gen 3 locking block, which Beretta had to pitch with the A3 instead of a sweeping revision because of data package bureaucracy. I expect the max round count is around 20k real life rounds with the gen 3. When a clone can justify its deviations from a 3 million round count multi-environment study I'll say it's probably as good or even possibly better, but that data being absent is hence why people feel compelled to buy the real McCoy. If it were easy everyone would do it for the lower price, they can't because it's not. If Beretta, one of the oldest firearms manufacturers in existence, can't do it better for cheaper then it's a safe bet nobody can.
While some improvements may have been made, it is unknown what the objective of the changes are. I can add a lighter barrel to a precision rifle and call it an improvement for hunting, but I'm not fixing a flaw. I can put a lightweight hammer, recoil spring, and match trigger bar on the 92, but it's a less reliable weapon with those improvements. In the end, it would debase the rationale for purchasing into the platform's strengths. People buy the M9 because they want a TDP weapon that lives up to its role, a clone will never have that attribute— especially acquiesced through lower price point imitation. As far as a Norinco 1911 vs a Colt, I can almost say for sure that metallurgically the Norinco is an inferior product. Does it matter in reality, I'd say probably so when it's my own money getting put down for it. I'm not a firearms designer, so I hazzard on the side of better safe than sorry. You can find clones that are viable, but you run a risk. It's entirely more likely someone will tell you what you want to hear just to separate you as a consumer from your money than it is that someone found a way to undercut the originator of the product.
