Nonradiative simulations that only include heating due to gravitational processes fail to match the observed mean X-ray properties of galaxy clusters. As a result, there has recently been increased interest in models in which either radiative or entropy injection (and/or redistribution) plays a central role in mediating the thermal and spatial properties of the intracluster medium. Both sets of models can account for the mean global properties of clusters. Radiative alone, however, results in fractions of cold/cooled baryons in excess of observationally established limits. On the other hand, the simplest entropy-injection models, by design, do not treat the cooling structure present in many clusters and cannot account for declining entropy profiles toward cluster centers revealed by recent high-resolution X-ray observations. We consider models that marry radiative with entropy injection, and confront model predictions for the global and structural properties of massive clusters with the latest X-ray data. The models successfully and simultaneously reproduce the observed luminosity-temperature (L-T) and luminosity-mass (L-M) relations, yield detailed entropy, surface brightness, and temperature profiles in excellent agreement with observations, and predict a cooled gas fraction that is consistent with observational constraints. More interestingly, the model provides a possible explanation for the significant intrinsic scatter present in the L-T and L-M relations. The model also offers a natural way of distinguishing between clusters classically identified as cooling clusters and the relaxed non-cooling clusters. The former correspond to systems that experienced only mild levels (300 keV cm2) of entropy injection, while the latter are identified as systems that had much higher entropy injection. The dividing line in entropy injection between the two categories corresponds roughly to the threshold for massive clusters. This finding suggests that entropy injection may be an important, if not the primary, factor in determining the class a particular cluster will belong to. These results also suggest that the previously identified relationship between inferred flow strength and the dispersion in the L-T relation is a manifestation of the distribution of cluster entropy-injection levels. This is borne out by the entropy profiles derived from Chandra and XMM-Newton. Finally, the model predicts a relationship between a cluster's central entropy and its core radius, the existence of which we confirm in the observational data.