Yes, you can, and should, model this as two separate diaphragms. At 5' separation you could model this on 2 distinct stories or model it on one story/level type and then adjust the column and wall elevations to raise or lower one of the areas by 5'. For overturning moment summation, the elevation of the force is derived directly from the original story data, so I recommend starting with a story height either between the two levels or at the higher elevation to be most conservative.
Drift results, etc are based on two separate diaphragms, though they could be linked with lateral beams if you wanted.
Wind calculations are based on two fully exposed towers. Where that assumption isn't perfect, user defined wind loads are my recommendation.
http://communities.bentley.com/Products/Structural...

@SethGuthrie Thank you so much for your quick response. Just to confirm that the drifts are computed at center of mass? and when there are two diaphragms, RAM separate them while reporting. Also can you help me to find out some verification examples for lateral analysis.
I am user of ETABS for last 10 years and now starting use of RAM and therefore I have lots of question.

With drift, you can get results at the center of mass of each diaphragm (Reports - Story displacement) or for any graphically selected point (Process - Results - Drift at a point). It should give you what you are after.
Our community site has only a little bit of hand-verification samples, mostly for the design, not the analysis. http://communities.bentley.com/other/old_site_memb...
I don't have anything more general to share (regretfully the internal test benchmarks are not in a state to share).