Sorry, juzz, but we've gone over this one so many times on this Board that it's starting to get a little, well ... I'll explain this whole thing.
First of all, all countries base their production performances off the final version of a series of special 'prototypes'. They're a progression of prototypes that advance through stages until the designers/gov't. contractors are satisfied - or not. That final prototype is set to the same weight and configurations as the envisioned production model. Thus, theoretically, there really shouldn't be much of a performance difference between the final prototype and subsequent production models.
In practice, this is not the case, for a number of reasons. The most obvious are differences in assembly methods: a single prototype is tended to by the top engineers and technicians of the company with comprehensive testing over that single aircraft, while production models are cranked out at assembly plants, and shipped off to their units after passing minimum requirements. So, as it turns out prototype data is generally superior to production data.
The problem is that the Soviets kept records of their production data for quality assurance reasons (moving their industry to the Urals created many problems in quality control), something not generally done. This data is usually stated along with the prototype data nowadays, creating the impression that only the Soviets had such discrepencies. This is not the case, as all countries had similar differences between the two, it's just that the Soviets have it published while the most other nations do not.
