Right, so even if I put the loading ("Transfer Rate" and "Write %") at 0, changing the clock speed from 800 to 400 still yields ~800 mW reduction (at 35C case temperature), which I do not see.
This leads me to believe that either my initial power is lower than than calculated by the spreadsheet (maybe due to enabled power saving features), or I have not fully accomplished the clock reduction.
I will instrument my 1.5V rail and get back to you.
Thanks for the quick replies.