On a serious note: It appears that the thrust vectoring servos are simply too slow. Linear Control Theory has been pretty well understood since the mid 1960s. Any system can be driven into instability via excessive gain (over reaction) or by excessive lag time. Slow servos act like excessive lag time.
Slow servos could be an attempt to use a legacy part. Slow servos could be driven by a packaging issue. There might not be enough room to package a bigger electric motor or a gear-motor that allows the motor to spin at max power. Or the design might not have made accommodations for motor cooling. You can get a huge amount of power out of a very small package if you can vent some liquid N to it for cooling.
It is clear the Russians have a problem. This is not the first Proton rocket to fail at launch.
What is not clear is if the Russians will force transparency into the problem solving and make fixing the problem more of a priority than CYA.