Fuzzy logic has emerged as an efficient strategy to develop accurate nonlinear controllers. This work presents a novel design methodology of dual-input single-output fuzzy logic controllers in search of a cost-effective solution. For this, instead of synthesizing the classical three-stage fuzzification, rule inference and defuzzification units, a further stage is added inspired in modeling the resultant control surface: the whole nonlinear area is split in specific rectangular sectors and each of them is approximated by a second-order polynomial function. In this way, the hardware-software co-design of the fuzzy controller is reduced to a computer that for each input point (x,y) calculates its output z=f(x,y) according to the surface model of the particular sector that encloses this input point. The physical implementation of the fuzzy controller is based on a MCU-FPGA platform where the control surface is segmented, parameterized, stored through initialized SRAM memory and appended to the design bitstream as a simple and customizable data file. Moreover, the dynamically reconfigurable FPGA makes feasible to multiplex the silicon-based functionality at run-time while the sequential execution of the fuzzy algorithm is in progress.