A clear answer is not available, since no one performed a proper test as of June 2021.
The closest answer I can provide is that damping feet result in a change of vibration behaviour, as discussed in this Klipper issue report. How it's not clear.
I paste some of the findings:
I took several measurements with different configurations and the
results are quite surprising IMO. The following graphs show the
respective vibration results for (from left to right):
1- The retrofit back to the standard hard rubber feet
2- The printer
only standing on the "feet clamps" as shown in the picture above
The printer standing on the "feet clamps" but with squash balls
Configuration 2 vs 3:
The damping is causing an additional spike at ~76 Hz;
Power-Spectral-Density is roughly the same;
Overall vibration is a bit lower in the damped version (11.4 % vs 16% @ 3hump_ei) --> Probably due to damping the additional spike;
A bit surprisingly the shaper recommendation is identical
Configuration 1 (original feet) drives Power-Spectral-Density from e3
range to e4 range
Configuration 1 brought back a dominant spike at
around 75 Hz to 80 Hz (which was surprisingly gone with the printer
sitting on the hard and small area of the "feet clamps"
1 nearly tripled the spike at 75 Hz to 80 Hz compared to No. 3
(Power-Spectral-Density 12,000 vs 4,200) The spike at around 30 Hz
stayed pretty much the same
The effect on the Y-axis is pretty small
The resonance effects very much depend on the type of feet
whatever reason every damping in the feet resulted in worse vibration
If you want to know more about it, install Klipper and an ADXL345 accelerometer, make measurements, write a blog post or post here in a new answer.
I may be able to do it within within the year. It's a loose deadline, yes.
I have a printer on a wobbly table, but I haven't performed measurements yet.
I expect it to be better than a stiff table, since part of the energy is dissipated by the table and not by the printer frame (we are not talking about independent feet, but table as a whole), resulting in better quality at high acceleration. Of course if your printer runs at 500-1500 mm/s^2 you won't see anything. If you run at 5000+ mm/s^2, then...