We analyze some basic issues associated with Generalized Poisson-Kac (GPK) stochastic processes, starting from the extended notion of the Markovian condition. The extended Markovian nature of GPK processes is established, and the implications of this property derived: the associated adjoint formalism for GPK processes is developed essentially in an analogous way as for the Fokker-Planck operator associated with Langevin equations driven by Wiener processes. Subsequently, the regularity of trajectories is addressed: the occurrence of fractality in the realizations of GPK is a long-term emergent property, and its implication in thermodynamics is discussed. The concept of completeness in the stochastic description of GPK is also introduced. Finally, some observations on the role of correlation properties of noise sources and their influence on the dynamic properties of transport phenomena are addressed, using a Wiener model for comparison. implies in turn that the higher-dimensional Cattaneo equation does not preserve non-negativity [49], which is highly unpleasant when this equation is applied to describe the space-time evolution of molecular concentrations, absolute temperature fields, or probability density functions, which by definition attain strictly non-negative values.Recently, we have started a systematic study of Poisson-Kac processes and of their higher-dimensional extensions (GPK) with the fundamental goals just mentioned above [50,51,52,53,54]. In developing such a research program, one encounters some delicate issues related to the very basic properties of Poisson-Kac and GPK processes, that require a careful dissection in order to avoid misunderstanding and misinterpretations.One of this issues is the Markovian nature of these processes. This and related issues are analyzed in the present article. In point of fact, starting from the understanding of general concepts, new formalisms and novel results are developed.This article is organized as follows. Section 2 briefly reviews GPK processes and their properties that are useful in the remainder. Section 3 discusses their Markovian nature, and from the extended Markovian condition, the adjoint formalism for GPK processes is developed, tracking the analogy with the classical Kolmogorov theory of forward and backward Fokker-Planck equations. Section 4 discusses the emerging fractality of these processes versus their local (short-term regularity), analyzing some of their implications in stochastic energetics (Section 5).Closely, related to these issues is the concept of completeness of stochastic description, addressed in Section 6. Finally, 7 discussed the role of correlations in Poisson-Kac processes. Comparing a Poisson-Kac dynamics with the evolution of a Langevin equation driven by Wiener fluctuations, possessing the same exponentially decaying correlation function, it is shown that the two model do possess radically different qualitative properties. This observation can be further elaborated in order to describe qualitatively different tunneling pheno...