I certainly agree with the question. Assuming that the D7k meets the expectations Nikon has set for it, it really is going to set the cat amongst the pigeons. It was already true that the D90 was able to handle the vast majority of the low-light photographic problems, and now the D7k improves on that significantly.
Whatever succeeds the D700 will be better yet than the D7k - it has to, since Nikon has to maintain the competitiveness in its top-line models, of which the D700 is a derivative. However, I think we're reaching the point of diminishing returns. I'd expect a D700++ to match the D3s, if not literally using its sensor. And that'll be fine.
But at, say, 50mp in FX or about 24mp in DX we probably will reach the limits of optics, printers and the human perception system. I'd say that the inevitable progress in silicon fab will yield some surprising turns. For example, we may get what could be called an 80mp sensor, but with only 25mp output. Instead, triangles of sensors might be used to "vote" on the value of a pixel - and that might yield ridiculously noiseless output, since the three pixels should all be the same. If they're not, or aren't in some identifiable pattern, the differences would be noise and eliminated.
Wild stuff like that could be under development already.
_____ Brian... a bicoastal Nikonian and Team Member
My gallery is online. Comments and critique welcomed any time!