For Nvidia, they used drivers released in November that had specific-Far Cry 4 support.
For AMD, they used drivers (Catalyst 14.9) released in SEPTEMBER, that did NOT have specific-Far Cry 4 support.
There was a beta version of the driver released at the same time as the Nvidia driver. There was an official version available December 8th. Both offered vast (up to 50%) increases in performance vs previous drivers.
I realize that they have to go with what is available during testing and you may hesitate to use beta drivers. However, December 8th seems reasonable when the article was published on December 30th. By using the September drivers, you significantly biased your review toward Nvidia. It is essentially worthless in current performance comparison. Essentially, Nvidia was given the opportunity to get better scores and AMD was not.
They should do an updated review with current, real-world results.