remember folks..... we are talking about a FF camera with the same pixel size as the 7D2. It will have very similar IQ to the 7D2. It will have very similar ISO performance to the 7D2.
You can't have the megapixel count of tiny pixels and the performance of large pixels. You can have one or the other... or a compromise. You can't have both. The 6D, 5D3, and 1DX will be superior in low light.
That's why I think it will be a new series, not an update of an existing series.
I disagree. The D800/D810 both demonstrate that similar-to-better IQ can be had with smaller pixels. Hell, the constant increase in megapixels in APS-C sensors for the last decade have been proving that for some time...every successive generation of APS-C sensors, each with smaller pixels, has produced better IQ than the preceeding generations, that includes Canon sensors, Nikon sensors, Sony sensors. Same goes for FF. We have observed a progressive reduction in pixel size while concurrently observing significant increases in overall image quality. So, yeah, I disagree. We CAN have both. We HAVE had both before...just not from Canon.
A 50mp FF from Canon will have much better IQ than the 7D II. Pixel size is immaterial, sensor size and sensitivity (Q.E.) is what matters. On a normalized basis, such a camera should crush the 7D II in terms of IQ, because your averaging together all those tiny pixels to produce the same output resolution. That's the same as having bigger pixels. Larger image downsampled means a reduction in noise, therefor an increase in SNR. No different really than stacking a bunch of frames together to reduce noise and increase SNR. Or simply running one of dozens of varieties of noise reduction algorithms on a single unscaled frame. It all increases SNR.
Total light gathering capacity. I don't know how many times I've said that on these forums, or how many more times I'll have to say it...but that's all that ultimately matters for IQ (terrestrial photography...astro is a little different).
I'd also go so far as to say that this sensor, if it has the higher Q.E., would also be superior in low light to the others at best, and no worse at worst. A loss in fill factor should be overcome by the increase in Q.E., or at worst, would simply balance things out (so high ISO performance shouldn't necessarily be worse.) A BSI design would eliminate the fill factor issue (and Canon does have some patents for BSI), and I'd say a 50mp BSI FF sensor should have BETTER high ISO IQ (since you lose significantly less light with BSI than any FSI design, regardless of pixel size.)
Assuming this 50mp behemoth has the same Q.E. (59%) as the 7D II, then it will have the same PER-PIXEL IQ (unnormalized IQ), but higher overall IMAGE IQ on a normalized basis. Same subject, same framing, same output magnification
== FF 50mp kicks 7D II ass.
At least for sensor IQ. For action, we'd still need a high frame rate and fantabulous AF system to achieve the same kind of subject freezing power as the 7D II. That's probably unlikely.