I truly believe (or hope? ) that there will always, for the foreseeable future, be a mission for both a high density FX camera and a more agile lower density FX sensor like the D700, although the resolution of future D700's will surely creep up to satisfy the market's craving for pixels. I don't see the D700 as the "old shoe" that I suspect high end DX is headed for.
I think we have to assume, for the sake of speculation, that there will never be a day where bandwidth is totally unlimited. Therefore, as you say, there will "always" be a struggle between resolution and frame rate.
And it appears that, at the moment, sensor density is proceeding faster than the ability to process all those pixels. But that does not preclude an FX camera of similar density from achieving the same frame rate as the equivilent DX camera and that is the key, I think, to this "future of DX" debate..
In five years or so we can come back to this thread and see if I was remotely close
There is a potential flaw in my arguments. My arguments assume there is some finite end game to the ever-increasing sensor densities such that there will truly be "equal film stock" among sensors..
The D800 is "only" 15 mpx DX, yet even before the initial shipments were completed Nikon annouced 24 mpx DX. So now we need a 56 mpx FX camera to make my argument of "same film stock" fully valid in the digital world.
I thought 16 mpx was the practical limit, yet I was proven very wrong by the D3200 (and the Sony camera of course). Therefore, I certainly am not going out on a limb to predict that 24 mpx DX "is it". Maybe someone else has better insight on that.
As long as each generation of DX cameras gets denser sensor resolution then there will always be this "arms race" between FX and DX where FX will probably lag because it in particular will be closer to the electrical bandwidth wall simply because it is passing 2.3x as much data. (Otherwise it could be argued that the D800 could have been 56 mpx if Nikon so desired!)
From that I concluded that in order to confidently predict the end of high end DX you have to first predict the end of the megapixel arms race