Option B looks at individual energy efficiency measures (EEMs) in isolation. Option D looks at the whole building. Option B is better for retrofits, where perhaps only a few EEMs have been implemented. Option D is usually better for a new building, where many of the EEMs might interact.
Many EEMs interact. What does that mean? Let's look at an example.
If I reduce the lighting power density in my project, I save electricity that would otherwise be used by the lights. But because lights also emit heat, I also affect my heating and cooling loads. To understand the full impact, I need to look at all these loads. Option D is really the only way to fully assess the performance of an "interactive" EEM.