The file is there for download, and if you change to Process 2012 you will see a slight reduction, but when adjusted to the final look of the image it is very close between Proccess 2010 and Process 2012.
(I happen to prefer 2010 because it gives a more classic look and allow much more lifting of shadow details. I find the Process 2012 too artificial looking).
It's the lifting of the shadows and the adjustment of exposure that create the more noise.
I personally don't see it as a problem but many have been stirring at that pattern.
I am more concerned what the actual b&w difference is between M240 and M246.
I downloaded it and experimented with it for a bit in LR6 before commenting. I disagree. I feel the image is about a third to half stop underexposed, considering the extremely elevated ISO setting and its reduced dynamic range. I wouldn't shoot at 12500 ISO if I was looking for 1600 ISO tonality. Process 2010 just makes it look worse, IMO. And remember that LR6 does not have an optimized calibration curve for the MM246 yet.
It's a subjective thing, however—like one person might prefer Rodinal 1:100 vs another person's HC-110 dilution F. As I said, however, printed to an 11x17, no noise is evident when I look at the image as a photograph rather than with a magnifying glass. Fixating more on what the 1:1 pixel display looks like than on what a sensibly sized print (or realistically sized monitor display for web use) looks like is, to me, counter-productive. It's getting lost in the technology rather than enjoying the photography.
The difference between M240 rendered to BW and MM246 is the interesting thing, I agree. At/near base ISO, the differences are small; at these insane sensitivities, they're two very different animals from what I've seen so far.
I eagerly await the arrival of my MM246.
G