From what I've been able to learn over the years, there were only severe and intractable problems with Westinghouse main generators in the Fairbanks-Morse CPA-24-5 locomotives, which were variously models 498A, 498B and 498BZ, and mostly on the NYC. The same generator used in the 2000 HP units has not been the source of much comment, and the similar 497 in the 1600 HP C-line units the source of none I've found. You also will not find adverse comment about the 472 or 474 generators used in various earlier locomotives rated 1500 to 2000 HP.
The main problem with the application of the 498 in the 2400 HP C-line units was shifting of the field causing flashover during rapid wheelslip transients. This fault was said to have been enabled by Westinghouse having certified the 498 to operate at a higher voltage rating (thus kilowatt rating) for application in the 2400 HP units than it did in the 2000 HP units. Operation of single 2400 HP units hauling 14 to 16 cars is documented on film on the NYC, which itself could constitute enough overload to shift the generator field, and could have been made worse by the action of slip, quick derating and recovery causing a transient that could shift the field and short the commutator. It may also be true that premature wear of the commutator bars was happening as well, perhaps leading to their shorting, due to excessive heat generation. One might think both.
Of course, this same generator was later used by Fairbanks-Morse in the 2400 HP Train Master. In this locomotive, a more sophisticated control system was employed, which included transition (which is not normally found in Westinghouse-equipped locomotives) and multiple steps of field shunt. It also included a protective relay against generator flashover. I've never read of the kind of flashovers occurring in the Train Master which had previously occurred in the New York Central's CPA-24-5 units.
And, finally, regarding Westinghouse equipment, many will tell you that the generators and traction motors as employed in Baldwin locomotives were indestructible. Although the air throttle equipment left something to be desired as a standard fit, the other equipment (to paraphrase John Kirkland's book) was above average in performance.
The repowering of the New York Central's 2000 HP freight and 2400 HP passenger C-liners was not only due to dissatisfaction with the performance of the electrical equipment in the higher powered units, but also reportedly due to the NYC's dissatisfaction with the opposed piston engine specifically -- and one also notes the move on the NYC to correct the wild disparities of the early diesel purchases by re-engining some of everything with EMD engines, including the "Babyface" Baldwin units and the two BLW road switchers (all carrying the early 608SC) and at least one ALCO-GE passenger unit. There were also Fairbanks-Morse Erie-built units that were repowered, and there were others that retained the OP engine but at a reduced power setting to increase longevity. There was a much bigger picture here, a bigger thing going on -- the NYC wanted to chop off the peak of the high-cost-per-mile locomotive mountain it had built. Thus, labeling the repowering of the C-liners as a wholesale condemnation of the electrical equipment and its failing as regards the 498 in the CPA-24-5 is perhaps too generalistic.
Repowering is a constant topic on these forums, but seems in the final analysis to have produced as much mongrelization as it tried to eliminate. Many times, the total incompatibility of subsystems in different brands of locomotives has been mentioned, which is one factor leading to complication in repowering. The PRR Sharks, which led off this string, were, it would seem, a much more thorough rebuilding and repowering, which of course made them much more expensive. New locomotives were a better idea, and that's much of why the phenomenon died out.
-Will Davis