Based on the comments a few people have written in this forum since early in 2014, it sounds as though the core and shell electricity use must be broken out in the energy model and that the 15% "only applies to the CBECS method and is not used at all if you do an energy model."
(Of course the 15% is the minimum if the USF is too high.)

However, the LEED credit language still reads:
"The core and shell building’s electricity is defined as the electricity usage of the core and shell square footage, as defined by the Building Owners and Managers Association (BOMA) Standards, but not less than 15% of the building total gross square footage."
and then after that it breaks down in to the two options:
"Option 1. Determine baseline electricity use: Use the annual electricity consumption from the results of EA Credit 1: Optimize Energy Performance."

I can't find any LEED Interpretations or Addenda that say the GSF/USF equation doesn't apply to CS projects that do an energy model. So was this really changed? Or was it an oversight on the part of a LEED Reviewer / or LEED Support person?