Just like the 7-3 chart, not really useful except for the first 3 simple cases (sphere, cylinder, planar surface/width).
One can derive different datums from some of the features shown, and it won't make any difference - because a datum is just an intermediate step before establishing a DRF.
For example, usually a flat taper is defined by applying the concept of "common datum feature" (previously "multiple datum feature"). One face can be labeled B, the other C, and the feature referenced as B-C. Each face would be mated to a plane simulated by the fixture or a computerized process. These two planes and the line of their intersection can be considered the datum instead of the Line-in-Plane idea. It doesn't take too many skills to figure out how to relate such two planes and a line datum to two planes of the datum reference frame. Call that a "scribble" but sometimes you get multiple equally good ways to a solution. You should not always expect someone feeding you the only one way "allowed", be it Y14.5 or Tandler. At least Y14.5 doesn't try to tell you that the solutions shown in the 7-3 chart are the only types of datum you can ever use. They are presented as examples, contrary to Tandler's pretentious "The Six Possible Datums" from your preferred example.