Eh... i'm only about half way through this.
Gordon and David McAfee, making the point that a reviewer is likely to test i different part of the game which may produce a different result to your marketing material is a fair point and true, however, its skirting around the real issue, these tech journalists are calling you lairs, incompetent, deliberately misleading your costumers and even calling the hardware worse than the previous generation, that is not something to be taken lightly, as just one of those things, its nothing blase' its very serious.
What you should be doing is giving these people your exact testing methodology, the exact point where your game testing starts, the exact rout and exactly where it ends, if they follow that exactly there is no way they can result anything more than a margin of error deviation from your own testing, IE 5% of the delta in your results.
When they come to you and say "these are our results and they are really bad", you don't just "oh well never mind that's fine" at the very least you ask them to replicate your testing to verify there isn't a problem, you even get their testing methodology to verify at your end, at the very least you should make sure they are under no illusion that your results are accurate even if theirs are too, because any excuse to throw mud your way they will jump on it eagerly.
I hope you learned something.