The studies focused on the thermally pulsing asymptotic giant branch phase experienced by low- and intermediate-mass stars are extremely important in many astrophysical contexts. In particular, a detailed computation of their chemical yields is essential for several issues, ranging from the chemical evolution of galaxies to the mechanisms behind the formation of globular clusters. Among all the uncertainties affecting the theoretical modelling of this phase, and described in the literature, it remains to be fully clarified which results are severely affected by the use of inadequate low-temperature opacities, which are in most cases calculated on the basis of the original chemical composition of the stars, and do not consider the changes in the surface chemistry due to the occurrence of the third dredge-up and hot-bottom burning. Our investigation is aimed at investigating this point. By means of full evolutionary models including a new set of molecular opacities computed specifically with the æsopus tool, we highlight which stellar models, among those present in the literature, need a substantial revision, mainly in relation to the predicted chemical yields. The interplay among convection, hot-bottom burning and the low-temperature opacity treatment is also discussed.