All of this is completely true, and always was true with film as well. Depth of Field scales were worked out on the basis of an acceptable "circle of confusion" IIRC, which was fine in the 1930s when they were defined with the films of the day and relatively small-size standard prints.
Indeed, I believe that the standard criterion used for almost all DOF scales on lenses and traditional DOF chart is something like what will not be noticeably out of focus to reasonably sharp eyes when
- printed at 7"x5"
- viewed from 10", which is about the diagonal length of the print.
Since what counts most is the ratio between viewing distance and image size, or to get fancy the angular size of the image, I would summarize this as saying that:
traditional DOF guidelines try to ensure that everything within the stated range of distances will probably appear in focus to most viewers so long as they
view from a distance at least as great as the diagonal length of the uncropped (or minimally cropped) image.Problems arise when images are viewing more closely that that, whether it be close scrutiny of large prints, or by viewing only a crop from the full image size for which the DOF guidelines are computed. The latter happens big time when a small fraction of the image is enlarged to fill a computer's display,
or the camera's rear screen. It also happens to a smaller degree with the "sensor crop" of DMF sensors smaller than the film format on which the DOF scale is based, and the related higher enlargement factor needed to get a given size of print.
