Traditional bulk microphysics schemes assume fixed ice particle properties (e.g., density or mass-size relations) corresponding to specific types of ice, such as cloud ice, snow, and graupel. In contrast, the Predicted Particle Properties (P3) bulk scheme predicts the continuous evolution of several properties in time and space for a user-specified number (one of more) “free” ice-phase categories. Each of these categories can evolve and represent any type of ice-phase hydrometeor. Since its inception about 10 years ago, P3 has undergone several major developments, including 1) predicting the sixth moment of the ice particle size distribution (PSD) to evolve the spectral width (3-moment approach) and 2) predicting the liquid fraction of mixed-phase particles. In the original 3-moment ice version of P3, simplifying assumptions were made regarding most of the process rate formulations for predicting the sixth moment of the PSD. In particular, for several microphysical processes it was assumed that relative spectral dispersion of the ice PSD was unchanged, meaning that in effect those processes were treated by a two-moment approach. Recently, a complete three-moment treatment for all ice-phase microphysical processes was implemented into the P3 scheme. In this approach, changes in the ice PSD dispersion are calculated explicitly for all relevant microphysical processes. 3D simulations of deep convective storm cases will be presented, including comparison with observations, to illustrate impacts of the new, complete 3-moment treatment of ice and the predicted liquid fraction of mixed-phase particles. Finally, prospects for future development of P3 will be discussed, including using a mixed-phase Lagrangian particle-based scheme to inform bulk microphysical process rates in P3.