A bit delayed, but here are the results from a bit of testing I did with comparing refresh boards with one of the best first-gen boards Rampage VI Apex.
Same settings confirmed by measuring the CPU Vsense, power consumption and CPU temperature. VRM temperatures were measured by the on-board controller readings and confirmed to be reasonable using an IR thermometer, and are normalized to the measured ambient temperature which was varying 2-3*C over the days the tests took place.
VRM temperatures with fixed airflow
A single mid-RPM 120mm fan was providing airflow over the VRM area.
VRM temperatures in different scenarios
Stock is with the default cooling solution without any active airflow. Passive means I unplugged any built-in VRM fans.
Power consumption, loss and efficiency
Values were calculated by measuring voltage and current both at input and output.
Another thing that's important to check when doing this type of testing is that the CPU is performing the same in all cases and that no throttling is present which would show lower temperatures. On X299 systems the most difficult to account for is Pmax or "phantom throttling". It's recently been added to HWInfo so it's possible to detect if it happens on your system or not. Essentially it kicks in if your Vccin voltage is dropping below a certain level, which is deemed as a good indicator that the CPU is drawing too much power. Asus boards has the "VIN Tracker" option which gets rid of this completely when set to Disabled, if left Auto it will disable it when overclocking from BIOS.