That I remember RISA doesn't allow for inclined diaphragms, nor I see the convenience of using them for portal frame buildings. You need in any case to ensure the proper stability items are present in the model; I recently designed one of these (simpler than yours, without crane) with RISA and still plan to review it next weeks for a proper representation of the actual geometries is always tricky (mine's had some geometrical irregularities embedded in the roof making this more relevant).
As to your question I don't see the need of use of any diaphragm, but I see that the efficient design of these structures with any program would require the repeated practice in several models, and maybe several (final) models per structure. This way one would gleam the better of the program and own's accumen; but I have not had the luxury of recent repetition of this kind of problem.
In my view, since these structures can have not only bottom shoulder flanges, but apex bottom flanges as well in compression, even the flange bracing should be represented, but this turns almost impossible when attempting to retain the main members as frame (line) members. That would ask for representing the main members as plate elements and once there you may be loosing the sight of the member as such. and so the convenience of using more than 1 main model appears.
