OK how about this as a possible working hypothesis?
Tell me, did the engineer(s) check any of the connections outside of your house, and, more importantly, did he check the street cabinet?
Suppose that what you have is a bad or loose connection somewhere between you and the cabinet. When the connection is good, you are getting the full power delivered to the hub, but a passing vehicle shakes the cabinet a bit and the connection loosens, down goes the power level, up goes the error rate and you have a problem.
The original installer puts you on a tab in the cabinet which gives a good upstream power level but the downstream is a bit too high, so he fits a forward part attenuator to the line to reduce this.
All is well until the connection degrades in the cabinet (bit of an assumption but seems fair enough). With the attenuator in place, the power is now too low and you have problems.
Now the second engineer turns up, measures that the power levels are too low and decides to put the attenuator - all fair enough, until the connection gets shaken up again and makes a good contact. Without the attenuator the power levels are now way too high, it overdrives the circuitry in the hub and again introduces all the errors.
Rinse and repeat.
Anyhow, like I said, but if a shot in the dark but does explain what you are seeing. You’ll still need another visit unfortunately to check the levels and connections from the cab back.
John