Sharpness and resolution are not the same thing. That 10% measured MTF difference equates to a tiny bit of sharpening in camera or in post. It does not mean greater enlargement potential.
When deciding between the two I would open files from both in PS, side-by-side, and try to equalize sharpening. I also printed samples with and without sharpening on the 5Ds file. Here were my observations:
- While pixel peeping they were difficult to tell apart without looking at the file info or finding a section of the image with very high contrast detail. With low to medium contrast detail there was typically no difference apparent to the naked eye. (This was at 100% and 200% on a 218ppi 4k monitor.)
- USM of 20-30%/1px would make high contrast detail as sharp or sharper on the 5Ds. But if I did that then low contrast detail often became sharper on the 5Ds file.
- 5Dsr files, when pixel peeped, often show color aliasing and stronger stair stepping effects. As an example, you can find both in the dpreview studio scene in multiple places.
- In fairness to the 5Dsr the color aliasing and stair stepping would never show in print. But then again, neither would the sharpness difference. Whether I sharpened the 5Ds file to match the 5Dsr file or not, I could not tell them apart at 24x36". (Epson 3880 and Hot Press Bright paper, one of the highest resolving combos you can find.) I could only find differences while pixel peeping.
Once I realized that the AA filter simply made no difference at normal print and view sizes, and was smaller than any of my other choices (lens; aperture; sharpening in post; paper) in the chain even while pixel peeping, I went for moire resistance.