There are many reasons why this improved power supply should result in better performance.
The first is simply locating a 50/60 Hz switching device away from the inside of the table where are located the sensors that supply feedback to actuators. The mains frequency and its harmonics are in critical zones of usual room resonances in audio systems, something that does not happen in labs!
As you said, the interaction with mains is always critical in switching or digital devices - probably the chokes are playing a double role in this power supply, also avoiding contamination of the mains with the intrinsic noise of the table electronics.
It should be noted that the bandwidth of this table extends bellow 1 Hz - and here rejection of slow mains fluctuations becomes critical. Also, these actuators need current pulses, and the power supply needs to supply them without any modulation that could affect the sensors. Probably it is why this circuit performs better than others for this task. Active tables are feedback devices - every small signal gets a lot of amplification, minimizing noise is critical. Curious people can read the Belleson regulator patent here:
https://www.google.com/patents/US8294440?cl=en