MOTM

I Fully agree with the idea of star grounding, and the ferrite 
beads, PCB ground plane etc. are also good for noise suppression.  
However, these synths are loaded with potential ground loops.  First 
you have two wires to ground on the power supply lines.  There 
probably isn't a problem here as both are closely paralleled.  
However, the modules are also grounded to each other via any and all 
patch cord grounds and also very possibly through the screws and 
rack rails, although the paint or powder coat may provide some 
insulation.  Ideally all patch cords should be grounded at one end 
only.  

Having said that, theory isn't the same as practice.  In wiring my 
studio I would only use patch cables with ground connected on only 
one end as a second last resort.  Often they make the situation 
worse instead of better and often you are trading less hum for more 
noise.  The last resort is to use one of those horribly unsafe gray 
AC ground adapters.  This is not a smart idea, especially if you 
have balanced line AC.

I still wonder about the MOVs though.  I'm no expert on these 
things.  But aren't they variable resistors which have a high 
resistance (impedance?) at low voltage levels but break down and 
conduct at high voltages?  If the breakdown point is current related 
then I can see how several in parallel would require a higher 
voltage before they conduct, but if the breakdown point is voltage 
based, the number in parallel shouldn't matter.  Also what is the 
relation to the speed the surge travels down the wire to the time it 
takes the MOV to breakdown?  Unless the breakdown is virtually 
instantaneous the surge is going to be traveling down the line to 
some degree so it would make sense to have MOVs at more than one 
location.