Hi,
I would say that interpolating to same size is a good way to compare cameras of different resolution. A better way may be to interpolate to a common resolution, say 180 PPI or 360 PPI for a given print size. The logic behind this that 360 PPI corresponds approximately to the resolution of human vision at 25 cm (10 inches), while 180 PPI corresponds to human vision at 50 cm (20 inches). Also, it seems that it is normally regarded that 180 PPI is needed for an excellent print.
There are more options to use Sony sensors than just the Sony cameras. Nikon, Pentax and others also use Sony sensor in some models.
Getting a Sony A# may be a decent idea, as it can take almost any lens. Prices on old A7-models can be quite low.
Personally, I plan to switch from Sony Alpha A-mount to Sony Alpha E-mount when Sony has a "pro" model. Zeiss now has a new 85/1.8 lens which I find highly interesting for the Sony, so I may switch earlier than expected.
Best regards
Erik
Slobodan, the comparison you posted has interpolations on the 5D3 and no interpolation on the D800 to try to normalize resolution between the 22 MP Canon and 36 MP Nikon. So the resolution part of the exam isn't really comparable.
However the shadow recovery comparison is enough to make any Canon fangirl cry
I have taken to over-exposing every shot that doesn't have outright highlight clipping (Michael's ETTR) - it helps some. I would love the Sony Exmor sensor in a Canon body, because I don't like dinky cameras, they don't feel right in the hand. However, I keep thinking about getting the upcoming Sony A7r II plus adapter, primarily for tripod use, where poor ergonomics don't matter..