Do specialists exhibit a greater efficiency and/ or rate of using their food resources for faster growth than generalists? For various reasons, this “feeding-specialization/ physiological-efficiency hypothesis” has had mixed support. Failure to reject the hypothesis may often be due to its complex definition (whether physiological, genetic, or ecological “trade-offs” are intended). Empirical tests of this concept with phytophagous insects have produced inconclusive evidence due to inconsistent nutritional quality in plants, genetic variation and phenotypic plasticity in the herbivores assayed, as well as the various types of “specialist/generalist” definitions employed. Complications for meaningfully comparing larval performances of monophagous and polyphagous species on different host (and non-host) species include constitutive (interspecific, inter- and intra-plant) differences as well as seasonal or induced variation in the allelochemical phytochemistry and leaf energy/nutritional value. Such experimental complications were largely circumvented by using concurrent bioassays with congeneric species groups of specialized and generalized Papilio that are both Lauraceae-feeders. Another complication was avoided by evaluating population and individual differences as well as insect species-level differences. Neonate larval survival of all populations of P. glaucus on spicebush across its geographic range was very poor (14% for 39 different populations) compared to P. troilus across all of its range (68% for 20 populations).  P. troilus on spicebush outperformed 23 species of Lepidoptera on many of their favorite host plant species, in controlled environments, including other family-specialized insects on Annonaceae, Aristolochiaceae, Simaroboucaceae, Pinaceae, Rutaceae, Magnoliaceae, as well as the multi-family feeders. General problems in measuring physiological costs as well as the broader significance of other nutrients and minerals in organism fitness, evolutionary divergence, and ecosystem stoichiometry are discussed.