thanks Wizzard; I don't see where I offended, but OP is not alone is asking the question why the different reviews offer such different outcomes; the test setup is crucial to answer this question;
in case of Crysis testing there are large differences because of OS, AA levels, Quality Settings, timedemo vs fraps vs gpu-flyover, system ram, 32 bit vs 64bit. Crysis gets a boost from more memory (7~8% 2gb>4gb) while high under Vista >> High XP;
Also CF scaling is not always the same under all circumstances, if you compare GTX numbers alone between reviews they are more in line (except crysis again OS/driver) , X2 is more different. I'm doing some more digging with testing X2 under XP vs Vista and AA scaling; Other sites have already done
some tests on the subject, but none really in-depth; digging through the 40+ reviews posted I focused on performance of Crysis, AA, OS used.
http://www.madshrimps.be/vbulletin/f22/crysis-aa-performance-ati-hd4870x2-vs-nvidia-gtx-280-a-56778/
like OP, no intention to offend; but OP calling out my review fake/manipulated why I have no stocks bought from either firm; and have to send back the tested products, there are no
favors here at all. blame it on Vista vs XP, timedemo vs FRAPS and last but not least... those reviews using Catalyst 8.7 vs Catalyst 8.8 beta; as the 8.8 beta bumped Crysis performance too.