The density theorem and Hausdorff inequality for packing measure in general metric spaces (Q1913501)

There are two basic mechanisms (diameter method and radius method) for extending the definition of Euclidean packing measure to general metric spaces. Although the diameter method has received the most attention in the literature, we show that it is generally inferior to the radius method in preserving the desirable properties of Euclidean packing measure. In particular, we show that the relationship with upper capacity, the inequality with Hausdorff measure, a version of the density theorem of Taylor and Tricot, as well as the pointwise representation of packing dimension with respect to a probability measure, are universally valid under the radius definition -- this is not so with diameter-based constructions. However, we also establish that the usual inequality with Hausdorff measure does hold for diameter-based packing measure provided the metric space contains at most countably many isolated points.