We present an analysis of 21 global MHD simulations of the dynamo action taking place in the interiors of fully convective M4 stars. We leverage our past work involving 45 simulations of shell-convecting-type M2 stars in a very similar parameter space to identify properties that their dynamos share and ways in which they differ. We find that the parameter scalings of many fundamental dynamo measures including the ratio of magnetic to kinetic energy and the cycle period are shared not only between the two sets of M-dwarf models but also with the broader literature on K and G dwarfs, suggesting that their convective dynamos are similar in nature. We find that the surface fields of the fully convective stars are stronger and more organized than those in the shell-convecting models and indicate faster spin-down than is observed, for which we suggest possible remediations. We note an abundance of dynamo states that are localized within one hemisphere at a time and find that such states are well described by a quasi-linear superposition of even- and odd-parity eigenmodes with similar amplitudes. We observe that the turbulent stirring of the deep fields of the fully convective models allows them to attain far higher peak amplitudes than those in the tachoclines of the partially convective models. We employ our AI-enabled pipeline, LoopNet, to search for buoyant, fibril magnetic fields in the simulated stars and find nearly twice as many such structures forming in the M4 models, in good agreement with observed differences in their flaring rates.