We give a detailed description of the differences between the factorization and results derived from SCET and QCDF for decays B → M1M2. This serves as a reply to the comment about our work B → M1M2: Factorization, charming penguins, strong phases, and polarization [1] made by the authors in [2]. We disagree with their criticisms.In [1] we derived a factorization formula for exclusive B decays to two light mesons using the soft collinear effective theory (SCET) [3]. Recently, Beneke, Buchalla, Neubert and Sachrajda posted a comment about our work [2], and compared it with their QCDF (QCD factorization) approach [4]. In this paper we compare results and reply to their comments [2].For easy reference, we summarize a few points made in Ref.[1] that disagree with Ref.[4]. We found that: i) a proper separation of scales Q 2 ≫ E π Λ ≫ Λ 2 in the factorization theorem are different in SCET and QCDF, and can be regarded as a formal disagreement if desired (Q = m b , E π ); ii) only a subset of α s (m b ) corrections are currently known, so results for these corrections are formally incomplete; iii) certain amplitudes are sensitive to the treatment of m c , with parametrically large ∼ v contributions from cc in the NRQCD region where v is the velocity power counting parameter; iv) current B → ππ data analyzed at LO in SCET supports values for the parameters ζ Bπ ∼ ζ Bπ J , in disagreement with numerical inputs adopted in QCDF. We also emphasized that Λ/m b power corrections need to be of natural size to support model independent phenomenology and verify that the expansion converges, concepts which are sometimes relaxed in QCDF phenomenological analyses. If power corrections change the LO values of ζ Bπ and ζ Bπ J substantially then this would indicate that the power expansion is not converging and a different expansion would be needed if model independent results are desired.We take this opportunity to also comment on results from Ref.[1] where we found agreement with points made in Ref. [4]. The original starting idea is the same, that factorization theorems for these decays should be derived by making a systematic expansion of QCD in Λ QCD /m b , where the terms in this expansion are model independent and unique. Earlier discussion of QCD based factorization methods for nonleptonic decays can be found in [5,6]. We agree on the scaling in Λ/m b for the LO amplitudes. There is agreement that input on nonperturbative functions can be obtained from B → π form factors and LO light-cone meson distribution functions φ π (x) and φ B (k + ). We also agree that at LO factorization occurs for amplitudes from light quark penguin loops, as well as tree, and color-suppressed diagrams. Finally, the set of the one-loop hard corrections computed in Ref. It is also worth emphasizing that the scope of our two works was different. In Refs. [4,8] factorization theorems were proposed based on the study of the IR singularities of lowest order diagrams in perturbation theory. Input parameters are taken from QCD sum rules. Certain power suppres...