I am not involved in marble research, but brief googling shows a lot of papers which used SEM for observation of marble.

It is a usual thing: different methods are used for different tasks. For example, for such routine task as metal grain observation/measurement optical microscopes are preferred (better contrast). For other things SEM is better. I believe there should be a lot of well-established routine jobs for marble with utilization of optical microscopes which do not need SEM involvement. For other things SEM could do better. Resolution alone is not a decisive factor in many cases.

If you obtain what you are looking for using SEM on marble sample than you should use it.

Hello i think that maybe this thesis could be usefull for your question.

http://d-nb.info/104621733X/34

If by "marble" you really mean marble, then optical microscopy can tell you a lot because you are looking at a polycrystalline material entirely made of calcite crystals, usually not too fine grained (in fact marble forms by a metamorphic process that favors coarsening of the crystals). Calcite happens to have an extremely high birefringence (in fact the highest among natural minerals), therefore any grain with the smallest difference in orientation is very well observed in a polarizing microscope by their different interference colors. So the texture of the marble is immediately visible at a glance with a relatively inexpensive technique. None of this is visible in a SEM and having a much greater resolution is of no help.

A different thing would be to look at the orientation distribution function by means of EBSD techniques. These are now automated and computer controlled in a properly equipped SEM. Whereas, to my knowledge, the same procedure is not standardized for optical microscope observations of interference colors. But a trained eye can go much further than a badly handlend sophisticated EBSD procedure.