Its momentary when you consider that graph and how much more time, by far, the speeds stay above reference.
30 minutes. 6 hours, six days.
Average all those clocks together and can you honestly say that the average would drop below advertised boost clocks at all? Of course you couldnt, because it would absolutely not. At least in this example you provided.
1) Why are you now only counting average clock speed? Follow the discussion, my post you originally quoted was in response to another poster. Maybe if you read that post it might make more since that what you are apparently arguing is not relevant.
2) Why are you still ignoring the computerbase.de results from within a case? I can't tell if you're arguing just to argue or if you're trying to set up straw men. Either way, it is irrelevant to what was the discussion at hand.