In regard to...
"At the risk of stating the obvious, doesn't it rather depend what you intend doing with the combination? I wouldn't want to try shooting birds handheld with a tiny body on a big lens."
Great discussion. My personal opinion is that the slr market is severely hindered by our pre-conceived notion that a camera should look like an slr. In the future, I'd anticipate this model to be broken. Right now, some people think of it as a shutter mechanism, as the name implies. Most people think of it as the big "camera-shaped" hunk of metal and plastic that makes it look professional-ish. I don't think the shape is at all optimized, however, for taking pictures, except for the use of very small lenses.
The discussion on mirrorless is a great example of why the term slr is increasingly inappropriate. If you stuck a mirrorless mechanism in the 1dx, would it not be an slr? And, on the other hand, if you took the existing 1dx shutter mechanism and installed it into a rifle-shaped mount that had balancing weights to adjust for different lenses, would it not be an slr?
20 years from now we may well have several different words we use for different types of rigs:
- slrs will have a different name and will be the category of cameras where you use both hands in front of your face. This will be the low-end tech, as the poor weight distribution alignment versus stability will be because the units lack a remote video feed to your eye.
- there will be several other categories of rigs that use different configurations:
- Rifle mounts; an interim design with better stability, but still needed only so long as direct eye placement is required.
- Remote stable platforms; which can be thought of as camera turrets linked to a photographer
- Wearable platforms; exploiting different parts of the body, usually close to center of gravity, with combinations of clothes, attachment mechanisms and new grip positions
- Direction-cognizant light collectors; where large surface areas of wearable material can collect light, and due to the direction of origin information being preserved, lens-like interpretations of that (massive amount of) light can be made in software.
- Insert your own fantasy here.
So for Canon/Nikon looking at the mirrorless trend, it's really an estimation of theirs as to whether or not that shutter mechanism happens to appear to be a long-term-worthy investment, versus a gear fad. On top of that, they need worry about it cannibalizing existing sales. Were I they, I'd attempt to design the best mirrorless mechanism, but use it to insert into existing constellation of form factors, and eventually new ones that would require the economy of space such mechanisms allow. There does not appear to me to be a good reason as to why mirrorless need be relegated to tiny sensors in the long-term. With optical viewfinders eventually being made useless by new form factors, the major disadvantage to mirrorless (no direct viewfinder) drops away.
It'll be an exciting development path. EOS-M could well replace the current Canon lineup of point-and-shoot cameras, which would be pretty smart on their part, because it opens up most numerous bit of market to lens purchases. It'll just take a few years for the brand managers in the various countries to chill out and let this cannibilization happen. Canon will be innovative in their patent portfolio, but let the market lead them, so they will appear to be fuddy duddy. -tig