Surely TOFSIMS data can be fitted and the peak position (m/z) can be precisely determined. It is not common in the literature because there's rarely a need for doing this. Peaks in TOFSIMS spectra are generally precisely located at the exact m/z positions with ultrasmall FWHM. Even two ions with very similar mass, for instance CN- vs C13CH, can be distinguished without curve fitting unambiguously. Therefore, curve fitting is almost only performed when calibrating the instrument.

XPS curve fitting is very common because they are useful for prying information such as chemical shift, which helps determining the chemical bonding in the analyzed material. Unlike sold state NMR, where peaks representing different bonding environments can be clearly resolved with good spectral resolution, XPS has limited spectral resolution(0.9~1.2 eV FHWM for lab source X-ray). Therefore, peak fitting is usually required to deconvolute the spectra and determining the peak positions of different components, which gives you bonding info. Peak fitting is also necessary for determining the area under the peaks, which is critical for doing quantitative elemental analysis. Therefore, curve fitting is required for both attributes of the technique.

no need

Whilst drivers for peak position accuracy are common to both techniques to provide answers to "what" is there, XPS is more quantitative than ToF SIMS in terms of peak intensities providing answers to "how much" of an element or moeity is on the surface. This is beacause SIMS suffers greatly from strong matrix effects with the consequence that standards are much more important in SIMS than XPS, where calculation can work better (e.g. using peak fitting techniques)