It pains me to say this, but I wouldn't be surprised if #1 is correct.
Sony axed the MVCD1000 (cdr 2.1MP, and my last camera) because it cost a great deal to manufacture, had more reliability issues and they generally made a loss on it. They replaced the CD1000 with the CD200 and CD300 for about $800 and $1000 repsectively at 2.1 and 3.? MP and presumably they are making money on them, but look at what you get for your kilobuck, they are a portable CDR drive with a rather pedestrian 'point 'n shoot' attached, nothing more.
Gone is the TTL LCD Viewfinder, gone is the Image Stabilization and rather tasty 10X zoom. Essentially, they have gutted all the things that we loved about the CD1000 bar the battery and the CD writing.
Mine died, and I bought a Pro90IS which I am almost obscenely happy with. I don't wish for higher resolution, I don't wish for anything much, bar perhaps a rev of the lens with less chromatic abberations in bright light at the extremes of it's zoom and maybe a more intelligent battery meter. The flash for the (all) Sony(s) is absolutely pathetic, really, pathetic (No TTL anything, put your finger over the magic eye and prepare for a shot of the white room, no focal length matching, tilt but no swivel, the list goes on). And it costs as much as my 420EX did. Grrr.
So, sorry, enough bitching about Sony, what can this tell us about Canon's next move?
Well, big lenses are expensive, as is IS. Designing new lenses is expensive (or so everyone keeps saying, though I'm not sure just how much I buy that. In the past, your Sinar or Hasselblad lens cost a lot in part because they had get a lot of very very clever people to do a lot of very hard maths to work out how to design one that would work well. That hard maths has not gone away, but it no longer has to be done by hand. Lenses (Or rather lens elements) aren't hard to emulate on a computer, so it should be perfectly possible to set up a simple CAD system that allows you to add or remove elements and see how different configurations behave pretty much in real time. Only send something off for prototyping when you are confident that it works well on paper, and lens systems are simple enough that when they work well on paper, (assuming you know the refractive index for your materials and so forth accurately) they work fairly well in practice as well (This comes from someone who designed and prototyped lens and prism mechanisms for industrial robotics, albeit terribly simple ones) Now, the engineering required to mount the glass in a robust and complex moving carriage is another matter entirely, that would indeed be a significant task, and how significant I do not know. Maybe this is what is meant when people refer to the expense of designing new lenses. Dunno, just skeptical) So, where were we,
big lenses are expensive, as is IS, Sony has given up (for the moment at least) on the superzooms and there isn't much other movement in that direction in the market place. To have a nice low noise high resolution image sensor with current tech, it has to be big, which means your superzoom has to be bigger. Which means more expensive and less appealing to the consumer, which probably means less sales, which is not good business sense. Now, if you can change one of those variables and put a high sensitivity, low noise small footprint CCD in the body of the Pro90 and tweak the optics so they will make meaningful use of it (which currently they wouldn't, in my humble opinion) then you have a successor to the Pro90. Otherwise you probably have variants on the theme of the G2 for a while, more MP like you suggest, and perhaps a point release of the 90 with new firmware and different decals and a couple of bells and whistles, or an F707 5MP 5Xish device (though I'd have thought we would have heard of the development of the lens by now, ala the Pro90)
Of course, now that I've said all of that, they will certainly release a Pro180 with 5.4MP, a 12X zoom, double IS and a coffee machine in the charger, and I'll look really silly, but either way I win.
It doesn't bother me a whole lot either way, if my Pro90 lasts like I hope it will, my next camera will be whatever the year after next's D30 is anyway. I've had enough with toy cameras, my wallet just hasn't caught up with my tastes yet. ;-) (Don't get me wrong, I love my Pro90 like a second liver, it is absolutely fabulous for an $800 camera and it gives me that special tingly feeling in my undies whenever I use it, but I grew up surrounded by Rollei, Hasselblad and Sinar so my tastes and expectations are somewhat misaligned with my current fiscal reality and I categorically DO NOT take photographs as my day job)
Right enough waffling, back to the Cold Fusion and if you read as far as here, I'm sure you won't risk another of my posts again. Sorry.
Dermot
Other Realistic Possibilities:
1: There will be NO new top end prosumer camera at PMA. Its too
soon for the G2 replacement and Canon has given up on the big zoom
due to money lost on the Pro90. Its going to concentrate on SLRs
for people who want big zoom. Note there was no top end Prosumer at
last years PMA either.
2: The G3 will show up. Essentially a G2 with 5mp.
Realistic can be boring.
like the currentG2 situation...