Not at all. With both type of loads you can know the moment at the shell and at the frame.
Only that you have to be consistent with the type of model you make (intermediate joints, rigidity of the shells, etc.) and be consequent with the results.
Typically, if shell is not important for you and you are just interested in frames, use "None" shells, without intermediate joints and uniform to frame loads.
A real application could be, e.g., a checkered plate supported on steel beams, where you do not really want to check the plate itself and you can neglect its contribution to the moment resistance of the beams.
On the opposite, you can find situations where you cannot ignore the shell.
For example, a concrete slab supported on concrete beams. In this case I would use shells with real rigidities, with a proper mesh so there would be (many) intermediate joints along the frames. In this case uniform load is the type of load to use. The moment at the beams and the slab is the one from a composite (beam+slab) section.
Of course there are many other intermediate situations where you should choose any possible combination of parameters...