I don't think there's room in the lineup for the 7D AND an entry-level FF DSLR AND the 70D. There's too much overlap between those markets.
When people write this stuff I always wonder what inside information they have. Because the only justification for saying this is if you know what the break-even point is for manufacturing any particular model of DSLR. Do they need to sell 10,000 units? 100,000 units? 1 million units? I don't know.
Well, I don't have any insider information but I do hang out with a few MBAs & marketing execs that also happen to be camera geeks, & when somebody releases something new we rap about it.
I agree that if a market analysis indicates that a profit could be turned on a particular model within the necessary period that a product would probably be greenlighted, but one thing that's clear whether we're talking about cameras or cars or computers: if you can hit two birds with one stone, do it. (also they all start with "c" <grin> ). They also have a certain number of price brackets in which they compete, and they have always clearly avoided having multiple DSLRs in the same bracket.
& I think that given the technology available, the models in the lineup due for a refresh, and
what the competition is up to for Photokina, the 7D successor and the entry-level FF camera will be the same camera, and that said camera will be very competent in fulfilling both roles.
I think "prosumer" users are ready to move to full-frame if the price is right. One need look no further than the number of threads on this forum à la "does this work on FF," "when I move to FF can I do X," etc. etc.
There's evidence that it's Canon's intention to do so as well, as all EF-S lenses released after the 7D came out are clearly "consumer" and not "prosumer". It doesn't make sense for them to make high-end or specialty/niche lenses that don't work on their best cameras (& unfortunately primes seem fall into the "specialty" category for Canon's marketdroids these days...).
As I mentioned earlier, the two things that have been holding them back until recently were:
- single digic fast enough to handle 8+ fps from a >= 18mp sensor
- the "reach" factor claimed by APS-C & APS-H telephoto users (a.k.a. pixel density)
The former has been fixed with digic5+, and so far in the small amount of time it's been shipping Canon's bet that the 1DX would appeal to APS-H shooters looks like it will bear fruit.
A baby 1DX for a grand less than a 5DIII would seal the deal. I'd certainly view it as a significant upgrade over my 40D, and the first camera to come along to warrant such a move that I'm actually willing to spend the money on.