Its hard to believe that INTEL the CPU giant at their 2018 INTEL power Supply Design Guide they mention Tektronix P6046 Differential Probe as preferable choice at measuring ripple.
Tektronix P6046 Differential Probe this consists of :
Differential probe + amplifier and power supply for the amplifier.
Both single probes they are placed in to a single handle, and I am assuming that they are matched at the factory.
The problem is the age of this probe, and the question of how accurate this can be today?
The use of two combined ( in the measurements menu) of oscilloscope Math and CH1 minus CH2 this will simulate a differential probe at a modern oscilloscope.
But if you combine the measurement error from its probe it self among with its one oscilloscope input, in the range of 5mV AC component, your are after of +/_ 5% of error.
There is a pile of Differential Probes in the market but all they are designed for higher voltages starting from 200V up to 6000V or higher.
The measurement of ripple at 3.3V up to 70V this is unique application, and a modern and accurate tool this is now missing from the market.