The D7100 is also an slr and no full frame - so that would be the combination of all draw backs. I would need SR stacking just to achieve the same resolution as I have with the 5D2. OK the DR would be larger. But I could just drop SR, and exposure-stack with ImageFuser.
You wanted more DR and more resolution than you can have with the 5Dmk2, and would also like less weight?
*The D7100 is 24 MP OLPF-less vs the 5Dmk2 21 MP (lense and technique probably is more important).
*The D7100 is reported at 13.72 EV of DR while the 5Dmk2 is reported at DR 11.86 EV of DR (both at 8 MP). Probably not the degree of improvement that you want.
*The D7100 weighs 675g vs 810g for the Canon
http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/865|0/(brand)/Nikon/(appareil2)/483|0/(brand2)/Canon
I think I'll need automatic alignments. Apart from the time it might take, I would not even know how to do what you describe. I'm not a software developer...
For now, I am in contact with Eugene (PhotoAcute) hoping that he figures it out. After all PhotoAcute claims to do this.
Thanks for your concern though.
I think that the problem you raise is interesting. "All" fancy (multi-)image processing applications want to have access to native raw files. And most of them are "one-trick-ponies". So what if you (for some reason) want combinations of HDR, SR, pano, focus stacking etc? And fit the files in your regular raw development workflow with profiles etc? Are not most of these tools about:
1) Fancy image _analysis_ (e.g. alignment)
2) Using this analysis to merge, stretch, mask etc a set of input images to form a more complex output image
3) Possibly fancy processing to make the complex image "fit" (crop, tonemapping)
-h