No the 828 is better than a dSLR in many ways and worse in many others. But since you aren't going to buy one and insist on interpolating your G3 experience to all other "prosumer" cameras, you will never know that.
All the V1 experience shows is that people will still buy the camera even if it's really noisy. The final production V1 shots are very noisy.
They look pretty bad to me too.
More sharpened, but not necessarily sharper. That will depend upon the lens.
You must shoot in very different conditions from me. I rarely go above ISO 100.
This is a good point that people haven't completley internalized yet. I did try reducing the images to 7x7 size to see what the noise looked like. It's hard to say without a directom comparison, but IMO they still looked pretty bad. Lots of areas that should be smooth still had a sort of mottled look to them.
From the 300D review, it appears that digicam manufacturers such as Sony are fibbing about the ISO rating by nearly two stops. (Otherwise the 300D is underestimating, but the end result is the same.) This suggests that for the same scene, you may not need to boost the ISO at all if the 828 is rated similarly to the V1.
A 2.8 lens has a 1 stop or less (at the long end) disadavantage over the 828 lens and this is before you factor in the ISO fibbing. The price you gave is more like the price of the IS version. The regular version can be obtained for around $1100, which is still a fairly frightening price. However, I find the weight even more frightening, which is the main reason why I haven't gotten on yet.
For me, noise was one of the main reasons for going DSLR. I take most of my shots at ISO 100 and love the noise free results. (I do have some expensive glass though...)
You're forgetting about Sony's tendency to fib about ISO. If they rate it the same as they rated the V1, the 300D will be nearly two stops more sensitive as the same ISO.
--
The 707 tested at ISO 100 was ISO 70 in real life. The 717 tested correctly (it exposes with higher shutter speeds than the 707 did in any given scene) and there is no reason to suspect the 828 will be any different.
--
There's no technical reason for producing incorrect ISO ratings for film. The standard (5800:1987) is extremely precise.
I'd rather not get into a long argument about testing standards here because even if we accept your premise, it doesn't change the fact the V1 is much, much less sensitive than the 300D acording to Phil's tests. The tests show much greater sensitivity for the 300D vs. the G5 at the same ISO, and greater sensitivity for the G5 vs. the v1 at the same ISO. this is true whether you believe the problem is fibbing or unclear standards.
The reason would be that the V1 has essentially the same size pixels as the 828. The 828 sensor is much more like the V1 sensor than the 717 sensor.
It's not clear that they're underestating it, but this isn't the point. The point is that your argument doesn't fly. It's partly off if sony rates the 828 like the 717, and way off if Sony rates the 828 sensor in a like manner to the most similar sensor in their lineup.
It sounds like you may not understand what dynamic range is. By definition, you can't increase noise w/o decreasing dynamic range - unless you simultaneously increase the capacity of the pixels to hold charge, but there's no evidence that this has occurred.
Identify the flaw in Phil's tests.
I never saw a test where he directly compared the 3 cameras for ISO.
By
MAC was complaining that the yarn was solid black. It was extremely easy to lighten the yarn and see a lot of detail without any particular amount of noise. This was due to a rather contrasty curve applied (probably by the original photographer choosing a "high contrast" setting).
