We compare assembly of dark matter (DM) halos with and without baryons from identical initial conditions, within the context of cosmological evolution in the ΛCDM WMAP3 Universe (baryons+DM, hereafter BDM model, and pure DM, PDM model). In representative PDM and BDM models, we find that baryons contribute decisively to the evolution of the central region, leading to an isothermal DM cusp, and thereafter to a flat DM density core—the result of heating by dynamical friction of the DM+baryon substructure during a quiescent evolution epoch. This process ablates the cold gas from an embedded disk, cutting the star formation rate by a factor of 10, and heats up the spheroidal gas and stellar components, triggering their expansion. The substructure is more resilient to the tidal disruption in the presence of baryons. The disk which formed from inside-out as gas-dominated is transformed into an intermediate Hubble type by z ~ 2 and to an early type by z ~ 0.5, based on its gas contents and spheroidal-to-disk stellar mass ratio. We find that only a relatively small ~20% fraction of DM particles in PDM and BDM models are bound within the radius of maximal circular velocity in the halo, slightly less so within halo characteristic radii—most of the DM particles perform larger radial excursions. The DM particles are unbound to the cusp region. We also find that the fraction of baryons within the halo virial radius somewhat increases during the major mergers and decreases during the minor mergers. The net effect appears to be negligible—an apparent result of our choice of feedback from stellar evolution. Furthermore, we find that the DM halos are only partially relaxed beyond their virialization. While the substructure is being tidally disrupted, mixing of its debris in the halo is not efficient and becomes even less so with z. The phase-space correlations (streamers) formed after z ~ 1 will survive largely to the present time—an important implication for embedded disk evolution.