I wouldn't call the difference in noise between the 1D IV and 70D a "wash". Compare the noise test results from DPR:
Why on God's green Earth would I compare RAW files with no NR when I never work that way and neither does anyone else?
If your photography consists of staring at zero NR, ISO 3200 black and gray patches at 100% until your eyes cross, buy a Nikon or a Sony. Then you can boost those RAW files +5 stops and really have some fun staring at patches.
Well, because the way YOU do things doesn't really matter for everyone else. However, for the sake of discussion, I most frequently do not apply NR to my images. I do everything in my power to maximize my signal strength up front, so that noise is minimal. I also usually publish most of my work with fairly considerable downsampling (my 5184px images are usually uploaded at 1140px on my blog), and the downsampling does enough basic averaging to reduce noise to an acceptable level in a standardized way (so if I downsampled and compared a 1D IV image with a 70D image after downsampling, the results would be the same.)
Noise reduction, in the case of RAW images, is not a feature of the camera. It is a feature of post-processing software. If we are to have an objective discussion about CAMERAS and their SENSORS, the ONLY way to compare the noise levels of two cameras is directly, with a RAW image. If you apply noise reduction, then export as a JPEG...well, we are no longer comparing RAW now are we? Were comparing denoised jpegs (and the algorithms that perform noise reduction.
Additionally, NR is a subjective process, and how well it works is affected by a number of factors. The noisier 70D images will require more noise reduction. The farther you push an NR routine, the more blurring it will introduce. The more blurring introduced, the lower the resolution of the 70D images, which puts it at a further disadvantage to the 1D IV. Comparing swatches that have been run through a noise reduction process means your no longer comparing the most objective data.
Sorry to burst your bubble here D by disproving your theory that the 70D has the same low noise as the 1D IV, but there it is in gray, black, and blue. I'm sorry that I have to share this kind of information, but to be frank, I honestly do not believe it is fair to anyone else to simply let anecdotal claims that have no basis in fact stand without providing as much objective and honest data as I can so everyone can see for themselves and make their own determinations.
The 70D is measurably and visibly noisier than the 1D IV. The data backs that conclusion.