Electron-positron momentum distributions measured by the coincidence Doppler broadening method can be used in the chemical analysis of the annihilation environment, typically a vacancy-impurity complex in a solid. In the present work, we study possibilities for a quantitative analysis, i.e., for distinguishing the average numbers of different atomic species around the defect. First-principles electronic structure calculations self-consistently determining electron and positron densities and ion positions are performed for vacancy-solute complexes in Al-Cu, Al-Mg-Cu, and Al-Mg-Cu-Ag alloys. The ensuing simulated coincidence Doppler broadening spectra are compared with measured ones for defect identification. A linear fitting procedure, which uses the spectra for positrons trapped at vacancies in pure constituent metals as components, has previously been employed to find the relative percentages of different atomic species around the vacancy [A. Somoza et al. Phys. Rev. B 65, 094107 (2002)]. We test the reliability of the procedure by the help of first-principles results for vacancy-solute complexes and vacancies in constituent metals.