NOT addressing the specific present icing condition -
One early turbofan (70s) had stall/surge problems in early revenue svc, associated with heavy rain. The factory had difficulty replicating the condition and the problem.
More field research brought forth the news the engine ran fine within the heavy rain, but when the a/c exited the rain the stall would occur.
So the test rig was modified to turn the H2O on and off very rapidly. Voila! the problem was now replicated!
Turns out the issue was the rapid air temperature change at the LPC discharge was the culprit; water in the LPC flowpath cooled that flow a lot, so the air temperature sensor at that point needed the be a faster-responding design to convey correct data to the engine control.
The water-ingestion tests during FAA certification were not aggressive enough!