Is it really worth the wait and expense when
cameras like the 40D can deliver awesome images, managable file
sizes, less weight and bulk and the choice of either standard and
EF-S lenses now?
Intrinsically, full-frame offers just one true advantage. At any given moment in the state of sensor technology, a full-frame sensor can collect more photons overall compared with a smaller sensor, including APS-C. This can be translated into more favorable tradeoff between noise at the pixel level and total pixel count (number of megapixels). As a result, the core justification for moving to full-frame comes down to whether or not APS-C can meet your own standards for both noise and resolution simultaneously.
So far we've talked about the intrinsic advantage. There are, however, several other advantages that arise in conjunction with the specific state of the photography market right now. That is, the specific lenses that are available for photographers to purchase and use (as opposed to theoretical ones that could be built but don't exist in anyone's catalog).
The first of these stems from available focal lengths. For any given focal length, an APS-C sensor will capture a narrower angle of view, requiring shorter focal lengths compared with full-frame to match AoV. If we could develop the existing range of lenses all over again, we could readily optimize for APS-C. However, the vast majorit of lenses in both the current catalog and in the used market were optimized for full-frame and thus don't necessarily provide optimal focal length. For example, my 24-135mm Tamron present a very useful range for full-frame but has too long a wide end on APS-C.
A related issue is depth of field. Once you match AoV between two cameras, one full-frame and one APS-C, you need to adjust the apertures differently as well to also match DoF. Specifically, the lens on the APS-C camera, already using a focal length that is 1.6x shorter compared with the one on the FF camera, needs to also use an aperture referenced by an f-number 1.6x smaller. If you do the math, you'll see that this is the same as both lenses having the same physical aperture diameter. Again, the existing catalog does not include lenses optimized for APS-C in this regard. You won't see a cheap 50mm f/1.13 lens designed for an APS-C camera designed to match AoV and DoF from an 80mm f/1.8 lens on full frame. It could be built, but the APS-C DSLR industry is too young to have revamped the entire catalog.
Another issue is viewfinder size and brightness. The focusing screen on an SLR is the same size as the actual sensor, since it sits the same optical distance from the lens (just along a different path while the primary mirror is down) and thus must receive the same projected image at the same focal distance. This means that you'd need more eyepiece magnification to produce the same viewfinder magnification. If you use a lens with the same maximum aperture (in f-stops) on both full-frame and APS-C, the APS-C camera's focusing screen will receive less total light (same light per unit area but less area) and thus will look dimmer when magnified to the same apparent size. To keep the viewfinder bright, APS-C cameras typically don't use this high an eyepiece mag and thus suffer from smaller viewfinders. Now, this could be remedied by the very same solution proposed for DoF ... using lenses with the same maximum effective physical aperture. That hypothetical 50mm f/1.13 on an APS-C camera would not only have the same DoF as an 80mm f/1.8 on full-frame, but its focusing screen would receive the same amount of light and thus could be magnified equally.
So, the FF advantage comes down to one intrinsic one and one that stems from available lenses. Of course, the story isn't at all gloomy for APS-C as long as you are already satisfied with the noise-resolution tradeoff, the focal length ranges, the shallowest DoF and the viewfinder size/brightness you are currently getting from today's APS-C cameras. For a great many people, the situation is good enough and not necessarily worth a minimum of an extra $1000 to change. This difference isn't likely to shrink for at least several more years since the increased size of a FF camera represents a significant extra manufacturing cost (an order of magnitude). You take a hit not only in the number of sensors you can make on a given wafer but in part yield as well.
APS-C happens to have an advantage of its own. Today's APS-C sensor have a tighter pixel pitch compared with FF sensors. This translates into mapping more pixels onto a given angle of view, which is good for getting maximum detail out of your telephoto shots. Yes, you could make a FF sensor with the same pitch. However, the resulting total pixel count would then be very high and would present a data flow issue for the camera (a need for high data bandwidth and buffer storage) that would increase the cost still further (and would likely decrease the frame rate, as with today's medium format backs).
David
I'm sure all the cameras you listed would look good from 10 feet away but there's times when you need and want to look at a print closer to see the detail.