Few problems with this.
First thing being, this is a biased AMD produced show. IIRC we saw things like this back when the FX was just barely out.
Secondly, as a reviewer there is no legitimacy at all. For instance, they don't show what they are encoding with Handbrake, nor do they show what video card is used on the game benchmarks. On that note, 1066 RAM being used which is slower than standard on the 980x vs 1600 on the dozer? Wrong answer.
Considering I get 14 points on a stock i5 750 for Fritz I highly doubt the 980x scores 24, or for that matter less than an 8150. The next point being, the FX 8 cores are only 4 cores with a physical iteration of HT.
If you look for Fritz benches it clearly shows the 990x (slightly faster 980x), 2600k, and 3960x stomping the 8150.
Cinebench performance of a 2500k to an FX utilizing 8 threads is only slightly slower. That should tell you something since it's 100% CPU utilization rendering. That's even on their own slide.
To wrap this up, the Bulldozer FX line has a slower single threaded performance than even Phenom 2 which was behind even the first Core i series in gaming. When dozer came out Sandy Bridge was already out dominating the CPU front and Piledrive is barely competing with SB against games priced slightly lower than the 3570k, which consuming more power and producing more heat than both SB and IB.
So in other words, the video and flame bait is rubbish. Benchmarks all the way around shows that the 8150 is slower than the Core i series for almost everything. The only competitive FX CPU in the whole line is the 8350 for its performance and price, but even then if you're a gamer the 3570k is a better buy. Haswell will release for almost the same price dropping the IB prices and putting even more of a tighter grip on the performance mainstream segment.
Edit: Total lol, I did that Fritz bench with Mass Effect running hogging half my CPU. Fritz is 15.17 without the game going.