In this thesis we evaluate and compare competing cosmological models for empirical and theoretical consistency and identify new ways of improving current paradigms of early universe cosmology. In the first part, we show that the most recent experimental data from the Planck2013 satellite measuring fluctuations in the cosmic microwave background favors a special class of “small-field plateau-like” models of inflation and disfavors the simplest inflationary potentials. We then identify a new kind of conceptual difficulty for the plateau models that we call the unlikeliness problem – namely, in an energy landscape that includes both plateau-like and simpler potential shapes, the plateau-like produces less inflation and, hence, is less likely to explain our observable universe. In addition, we show that the very same plateau-like models suffer from a new multiverse problem and a new initial conditions problem because they require that inflation starts at energy densities well below the Planck scale. Third, we comment on the impact of these results on the standard view of inflation and more recent versions of the theory invoking the multiverse and complex energy landscapes. In the second part of this thesis, imposing a single, simple, well-motivated constraint – scale-freeness – and using a general hydrodynamic analysis, we show that the unrestricted range of inflationary potentials reduces to a well-defined bundle of inflationary models. We classify and evaluate the scale-free inflationary models in light of Planck2013. We then repeat the construction to produce analogous scale-free bouncing cyclic models of the universe and compare with the inflationary results. In the third part, we introduce a new class of stable ekpyrotic/cyclic models that require less fine-tuning and generate negligible non-Gaussianity consistent with Planck2013 data.