Microcalorimeters are used by National Metrology Institutes (NMI) for the realization of the Radiofrequency and Microwaves (RF&MW) primary power standard. Since they are not available on the market, NMIs have to design their own systems. It involves the design of complex mechanical structures and of low noise data acquisition systems. Resulting setups allow precision measurements of RF&MW power by applying difficult and time consuming measurement techniques. Weaknesses to be addressed in order to improve the performance, whether in the mechanical structure or in the data acquisition electronics, are not easy to identify. Anyway, they can be found by analyzing the microcalorimeter measurement uncertainty. In recent years, several improvements have been made on the INRIM (Istituto Nazionale di Ricerca Metrologica, Italy) coaxial microcalorimeter with the aim of making its operation more stable and to reduce the measurement uncertainty. Its weak points are here analyzed by means of a comprehensive uncertainty budget. Details are also given on how uncertainty contributions have been determined. The method here applied is general and can be conveniently used in many other fields or applications provided that the analytical model is known. It is useful while designing an experiment and/or a posteriori, to identify the parts of the apparatus and of the measurement set-up that need to be optimized.