The algorithms for "finest detail" are optimised for vectors and for sharp edges and such. You'll likely see no issue using it for contones, but you won't see any benefit and the image will take more time to process because it will be so much larger.
You might actually see inferior results for some photographic images where soft areas are hardened or sharpened, though I haven't tested this specifically.
I don't think there is any sharpening involved, just the ability to maintain native sharpness by using finer detail when rendering the lines/edges. Sharpening vector graphics will cause jaggies to become visible.
Here is an explanation that makes sense, although there is still a benfit to printing at 720 PPI even when the image dimensions in pixels don't exceed 360 PPI at the intended output size. When using a good output resampling program, like several offered by Qimage, there is higher detail at interpolated edges/lines than the original data suggests. The interpolation algorithms can invent useful detail, which can subsequently be exploited at a higher PPI output setting, and one can sharpen at that higher resolution.
Cheers,
Bart