In short, all else being equal:
smaller pixels -> Less AA needed -> less GPU
more pixels -> More GPU and RAM
foveated rendering (eyetracking) -> more CPU, less GPU, ram
Next Gen is going to have more pixels and smaller pixels. If the resolution doubles (4x pixels), we'll need 4x the processing power if everything else stays the same, however, with eye tracking and foveated rendering you'll gain a factor of x2-x3 back ito performance under ideal situations, you will pay for this by using more CPU (to run the eye tracking and the foveated rendering stuff).
That said, I strongly doubt that even a doubling of resolution up from our current Rift/Vive standard will allow us to get flat panel fidelity in VR. You will still need AA to get rid of jaggies at 2k per eye. However, it will be immensely better than what we have today.
Now, DCS. DCS in general is heavily bottlenecked by CPU as it stands. Most of us today cannot get VR to run at 90 fps (except under special circumstances), even with a 1080ti which is the fastest GPU available.
Basically, unless ED does something about the CPU load (Vulcan?), next gen VR will only give us better resolution but poorer performance.