To me, the "elephant in the room" for diesels is the fact that almost all diesel operation uses either 2 or 3 multiple units, pushing the cost of a "consist" of freight diesels above steam. For those of us with layouts and the fact that we buy locomotives to OPERATE, the cost quickly gets prohibitive. Of course, modeling MODERN passenger diesels avoids this issue.
I believe the market for "named/famous" passenger trains is not saturated. Passenger trains might me successful if they do not weigh too much, they are prototypical in length, they have either warm or cool LED's depending on the era, they use insulated wheel sets and specifically NOT wipers, the diaphragms are not short lived, and they are painted absolutely correctly. I believe the brass 2-rail market for rolling stock for operators has largely vanished vs use of plastic. One additional important factor is that rolling stock, despite the homework required to do this correctly is basically lower tech than a hand crafted brass steamer or diesel. One impediment is the cost of plastic tooling.
I can think of two steam locomotives that MIGHT achieve the necessary ordering, although the price probably remains an issue.
The first would be a NYC L-2C, L-2D Mohawk. There were several variations of this locomotive that changed the appearance, including internal Coffin feedwater heaters vs sunken Elesco, water scoops vs none, and even two individual L-2D's with passenger pilots and one of these with Union Web Spoke drivers. The "as-built" versions also differed in appearance vs the late version with a larger sand dome and top boiler check. Of course the lettering also changed Antique Roman vs later block lettering. So several variations including the Big Four versions and the variations could be modeled. This model has never been made in O Scale/O gauge. All variations except the Union Web Spokes could be reproduced with minor changes in subassembly pieces.
The second would be a 2-6-6-2. An importer issued a NYC NE-2 2-6-6-2 mallet a number of years ago. They had driveline problems and were of questionable quality, and most of the few still in existence have either been modified or pending modification.
There should be a way to use the SAME 2-6-6-2 mechanism and with select use of subassembly pieces make credible 2-6-6-2's used by several railroads. The C&O's were the largest, but for MOST of these the driving wheel size was within a narrow range. That might imply use of the same mechanism. The boilers varied in size so this would have to be investigated. Again, the price point might be an issue.....