I will look for the letters I read the following information from however as Richard Hendrickson used to say at the start of his presentations "to put all this in context".
Context: The NP built 1000 boxcars for on-line needs in the first half of the fifties that once loaded promptly went off-line and stayed that way for several years. (Why you need to model NP boxcars). I would like to do a presentation/article on these as routing on five cars through two years was provided and those cars did get around the US and even into east and western Canada.
In the letters were a survey of the condition of the cars several years after building once a few came back and I do recall words to the effect that some cinder cutting is noted on the the roofing. The word some. No words that paint failure was noted to any large degree which would back up the observations of BC and SL, or that these cars needed to be shopped for repair/repainting.
Of roofs being sooty, I seem to recall several times Richard Hendrickson admonished us in presentations that during the steam era cars were sootier than many of us ever model. Photos of Richard's modeled cars back up that he modeled as he stated. Jack Delano photos of the cars seen in the downtown IC Chicago yard area by my prior recent postings underscored this also. And do not show severe flaking or failure of roof paint.
Please recall also that on most cars less than 20 years old - the railroad does not own these, the trust that funded the purchase of these cars does. And the railroad needs to keep up these cars to some minimal standards agreed to in the original contract that these were purchased under. Perhaps your studied railroad bought equipment outright with cash however for much of the history of what I have read, my studied railroad these car were purchase through bonds, hence the trust plates on the boxcars in builders photos. I'll see if I can find the paperwork that notes - cars painted under trust agreements - or something to those words. It is a sizeable group each year in shop records.
So I would petition that this negates to some degree extreme paint failure modeling.
At the end years of the STMFC stated time frame, and when some railroads were teetering financially, perhaps an argument can be made to differ.
Jim Dick - St. Paul, MN