In this contribution, I review the applications and potential limitations of the spectral energy distribution fitting tool that I have developed, with a strong emphasis on the limits to which this tool can be used to improve our understanding of massive star formation. I discuss why our current grid of models cannot be used to distinguish between the several competing theories of massive star formation. I also discuss stellar mass determinations, artificial correlations between parameters in the grid of models, multiplicity, confusion, dust assumptions, and unique fits. I briefly review the improvements we intend to carry out for our next grid of models, which will eliminate many of these limitations. Finally, I show examples of applications of this tool to massive young stars.