I recently did some tests by swapping round components in my two gaming systems. I currently have an FX8350 with a r9 290 and an A8-5600k with a hd 7870. Now my more powerful rig is plugged into a monitor where as my HTPC is plugged into a huge projector screen. I wanted to see whether it would be worth putting my r9 290 into my HTPC. So here are my results:
Unigine heaven: A8-5600k w/7870 @27fps, a8-5600 k w/290 @53fps, 8350 w/290 @54.1fps
Batman Arkham origins: A8 w/ 7870 @65fps, a8 w/290 @112fps, 8350 w/290 @103fps
F1 2012: a8 w/ 7870 @75fps, a8 w/290 @83fps, 8350 w/290 @91fps
Firestrike: a8 w/7870 @4709, a8 w/290 @7007, 8350 w/290 @ 8531
As you can see the bottleneck is small at the Most and in Batman Arkham Origins, the a8-5600k actually somehow beat the 8350 in the built in Benchmark. Both systems used exactly the same RAM (the same actual sticks) Hope this helps and people start to realise that bottlenecks are not as bad as people think. My results actually reflect those that were achieved by JayztwoCents on YouTube.
It would only bottleneck the the R9 290x if you play games that can take advantage of more than 2 cores. Most future games will still only make use of 2 cores. However, that is a general statement which may not apply to games that you intend to play which are using a game engine known for making use of more than 2 cores.
For example, Crysis 3 can make use of 4 cores. Therefore, technically any game that uses the CryEngine 3 can potentially also make use of 4 cores. Star Citizen (space combat simulator; 2015 expected retail release) uses CryEngine 3 and considering it is a pretty complex simulator using realistic Newtonian physics I am nearly positive it will use 4 cores.