Tom's Conclusion said:
The only way we can make AMD's FX-8350 look like a better gaming value than Intel's Core i7-3770K (specifically in the games and at the settings we used to test) is if the rest of the system is free. Because the rest of the system is never free, the FX-8350 never serves up better high-end gaming value.
This is the bottom line, just as guskline says. Getting an FX8350 and justifying the lower performance as being acceptable because of a price/performance basis only makes sense if you are somehow getting the rest of the system for free...which is the case when you doing a straight-up upgrade of just your processor (or at most just the CPU+inexpensive mobo).
As soon as the upgrade BoM starts to include DDR3 ram (maybe your existing rig is still DDR2 or not enough DDR3 ram), or an SSD (replace HDD or a larger SSD), or a GPU (replacing that HD5850 or GTX460), the value proposition of trying to save a few bucks with the FX8350 is thrown out the window. You will see yourself net better price/performance from the upgrade by going with the Intel processor.
The one major problem with this approach to the cost-benefit analysis is that this is not how most people approach upgrading or new system building. People don't generally log into Newegg and start buying the best of the best and figure the bank account will cover the tally, whatever it comes to.
Rather, most people generally approach their upgrading plans with a specific budget in mind. "I don't want to spend more than $600 on this upgrade" But they are wise enough to know they shouldn't strictly adhere to a rigorous limit because they may end up losing out on a good deal of performance if say they would have spent $615 instead of forcing their upgrade to fit inside an original $600 budget.
So the budget is bound but has some stretch to it.
And it is with this perspective in mind that price/performance analyses ought to be done. Anyone considering upgrading to either the 8350 or the 3770k is going to be doing so with a similar budget in mind. So they might get the 3770k but in doing so they must make a trade-off or two (no SSD, or less ram, or they take one step down on the class of GPU they are buying)...whereas if they opted for the FX8350 then their overall budget might allow them to buy that SSD, or extra 8GB of ram, or a video card that costs $130 more while delivering higher video performance.
So while I do agree with the conclusions drawn by the Tom's article, I feel they are flawed with respect to factoring in how most people approach price/performance upgrading and as such
the conclusions drawn by the article are likely irrelevant to the vast majority of people who are going to read the article.