Sure it'll have the power to plough through pretty much everything but that's alot of GPU power for a measly 4GB framebuffer per GPU for a dual GPU computational monster.
With the increase in video memory demands in some of the more demanding titles this could create issues if the games are running at their highest settings and higher than 1440p resolutions.
VR shouldn't be much of a worry though as none of the current VR headsets have displays that are higher than the combined 2160x1200 resolution, however I'm not sure of the memory demands of a VR setup.
The next generation GPUs are also launching this year with HBM2 and possibly GDDR5X, but they might not be able to put out performance close enough to a dual FIJI GPU.
Could it be possible and practical to revise the Dual FIJI GPU and add HBM2 to it? This would allow for more than 4GB memory per GPU. It could also be a factor behind the delays of this GPU.
AMD Dual-Fiji “Gemini” Video Card Delayed To 2016, Aligned With VR Headset Launches
The product schedule for Fiji Gemini had initially been aligned with consumer HMD availability, which had been scheduled for Q415 back in June. Due to some delays in overall VR ecosystem readiness, HMDs are now expected to be available to consumers by early Q216. To ensure the optimal VR experience, we’re adjusting the Fiji Gemini launch schedule to better align with the market.
DX12 is also believed to be able to combine memory pools of multi GPU setups, but we don't know how well this will work, how practical it will be and what developers will take advantage of this for their games. And also, what about the hundreds of DX11 games?
I don't mean to sound negative or anything, I'm honestly genuinely worried about this.
Look at the amazing longevity of the Radeon 7990 and the R9 295X2, they pretty much bested NVIDIA's competing dual GPUs.
The large framebuffer of 3GB for the 7990 was apart of this and is a factor which helped it beat the GTX 690.
The R9 295X2's pricing is what made it alot more affordable and feasible to consumers, as-well as it's 4GB frame buffer.
In all fairness NVIDIA didn't response really well to AMD's 295X2. They launched the TITAN Z which was a 12 GB dual GPU card with 6GB per GPU, but it's pricing was insane which made it completely uncompetitive with the 295X2 for gaming purposes. It has other uses, but in terms of gaming it was not a justifiable purchase at all.
Geforce - GeForce GTX TITAN Z
PCPer - The NVIDIA GeForce GTX TITAN Z Review
EDIT: I wrote way too much!
Edited by TopicClocker - 1/25/16 at 3:57pm