Right now the betting would be on Canon and Nikon, in that order. But a timely strategic announcement from Sony without one from Nikon could easily change that.
Why? Because people are more likely to move from brand to brand
than before. The very fact that Sony has any chance at all against
Nikon, to me, is living proof that it's not as hard as it used to
be to get people to switch brands when they buy a new camera.
Implicit in your argument is that buying a new camera every X years is going to continue at the same pace. In other words, everyone who buys a DSLR today will buy a new one in some fixed time frame in the future. However, since I have one of the larger databases of Nikon users in the world and can and do run data mining on it, I can tell you that I already see evidence that camera turnover has slowed. Indeed, it seems to be tracking similarly to what happened when the AF ruckus started in the film SLR world. For less than a decade, there was a fast turnover in bodies and low brand loyalty. Ten years on, the turnover was mostly gone and brand loyalty reappeared. Slowing turnover makes the current lack of brand loyalty an even more dangerous thing than it is. You can wait too long to tell your customers what your future will look like.
My statement is based firstly on anedotal evidence--many serious
photographers I know have sold gear online to finance new gear
purchases.
Many serious photographers sold gear to camera stores, through Shutterbug, and at local flea markets to finance new gear purchases. Nothing has changed other than the location of the sale.
Take the rate of change of film stocks and camera bodies
You mean like Velvia being the primary choice of most pros for almost 20 years, or the fact that pros like Galen Rowell still used an F4 four years after the F5 came out (and the F5 came out 8 years after the F4)?
and
multiply them together to predict the rate of change for digital
bodies, then add in the x-factor of innovation for capabilities
that are above and beyond what was possible with film. On top of
that consider the advances in manufacturing capabilities over the
past 20 years and what that means for how quickly features can
trickle down the line.
The assumption here is that you can continue to market that successfully (which doesn't bode well for a marketing fizzle like Nikon). Frankly, it isn't AF capability, megapixels, or dynamic range that are going to get increased automatic sales at the high end. The only two things you really hear 1DsII and D2x pros complain about enough that they'd immediately vote with their wallet are high ISO performance and WORKFLOW (and the former tends to be fixed by the latter). Giving us files that are 2x the size doesn't improve workflow, it hurts it.
Frankly, I'd rather have a D200-sized D2x than a 16mp D3x right now. And a Coolpix 8800-sized D200 instead of a 16mp D300. Both with better workflows.
I'm speaking vaguely, but hopefully conveying the idea--the rate of
change in SLR cameras is not going to slow down any time soon IMO.
Well, I agree with you there. It's been increasing and it will continue to increase. This is almost a given if you look back at any tech product maturation cycle. The next few years will be a continued frenzy of new product.
With a high rate of change, products get "out of date" more
quickly, driving repurchase.
Here's where I start disagreeing with you. The "driving repurchase" statement isn't consumer based, it's manufacturer based. In other words, the maker needs to drive the repurchase cycle but the consumer doesn't necessary need to accept it. In some markets we've seen regular update cycles work. The most notable example is automobiles, where the makers have created an expectation of new product every 12 months and taught the "average" consumer to dip into that once every three years or so (leasing is a good example of that teaching). But there are plenty of examples of markets that withered at the top of the growth curve. High fidelity audio is a good example, despite plenty of new features and an almost constant product updating.
Personally, I'm less inclined to believe that people will automatically update their DSLRs just because something newer comes along. You can already see some of that inertia in this forum, what with the "my D100 still performs" and "I upgraded from a D70 to a D200 but don't see any difference" posts. (Note: I'm not saying that I agree with the content of such posts ;
). I'll predict that this will get noisier and more vociferous with each new generation of DSLR.
--
Thom Hogan
author, Nikon Field Guide & Nikon Flash Guide
editor, Nikon DSLR Report
author, Complete Guides: D50, D70, D100, D200, D1 series, D2h, D2x, S2 Pro
http://www.bythom.com 
