A relatively simple and physically transparent model based on quantum percolation and dephasing is employed to construct a global phase diagram which encodes and unifies the critical physics of the quantum Hall, "two-dimensional metal-insulator", classical percolation and, to some extent, superconductor-insulator transitions. Using real space renormalization group techniques, crossover functions between critical points are calculated. The critical behavior around each fixed point is analyzed and some experimentally relevant puzzles are addressed. PACS numbers: 71.30.+h,73.43.Nq,74.20.Mn Two-dimensional phase transitions have been a focus of interest for many years, as they may be the paradigms of second order quantum phase transitions (QPTs). However, in spite of the abundance of experimental and theoretical information, there are still unresolved issues concerning their behavior at and near criticality, most simply exposed by the value of the critical exponent ν (describing the divergence of the correlation length at the transition). Below we discuss three examples which underscore these problems. First, in the integer quantum Hall (QH) effect, which is usually described within the single particle framework, various numerical studies yielded a critical exponent ν ≈ 2.35 [1], in agreement with heuristic arguments [2]. Some experiments indeed reported values around 2.4 [3], while other experiments reported exponents around 1.3 [4], close to the classical percolation exponent ν p = 4/3. Even more perplexing, some experiments claim that the width of the transition does not shrink to zero at zero temperature [5], in contradiction with the concept of a QPT. Additionally, the observation of a QH insulator [6] is inconsistent with the QPT scenario [7]. Consider secondly the superconductor (SC) -insulator transition (SIT), for which theoretical studies suggest several scenarios. Similar to the QH situation, some experiments yield a value ν ≈ 1.3 [8], not far from the value ν ≃ 1, predicted by numerical simulations within the random boson model, but closer to the classical percolation value. Other experiments, however, yield ν ≃ 2.8 [9], while some experiments claim an intermediate metallic phase [10]. As a third example, consider the recently claimed metal-insulator transition (MIT) [11]. The critical exponent is again close to 1.3 [12,13] but the occurrence of such phase transition is in clear contrast with the scaling theory of localization.The basic question which naturally arises is then whether it is possible to unify these two dimensional phase transitions within a single theory and thereby resolve some of the problems raised above. A hint toward an affirmative answer is gained by experimental indications that percolation plays a key role in both the QH transition [14] and in the SIT [15]. Moreover, its relevance to the MIT has been argued theoretically and observed experimentally [12,13,16]. The QH and the SIT have been treated within a percolation-like model in Ref.[17], consisting of SC or QH droplets connec...