I would advise you to remember that there are five factors I can see making this one very hard to keep working correctly:
1) Maintenance. Steam locomotives were maintenance intensive. They could last forever and didn't need major components replaced as frequently as diesels, but steam locomotive overhauls aren't easy work, and they require frequent maintenance to keep in operation with regards to fuel, lubrication and routine maintenance. Better components and standardized component design could fix a lot of these problems, but if you are going the path of buying secondhand locomotives from other railroads (the New York Central's Niagaras and Hudsons have already been mentioned here), you are adding to your maintenance costs. If you go this route and have a Andre Chapelon or LD Porta-like figure coming in to overhaul the fleet as you talked about, standardizing components would be a very, very wise decision. You also want to design everything possible with roller bearings, stokers, systems for ash removal and generally designing the improved locomotives for easier maintenance. Beyond this, the maintenance of many coaling stations, water towers, running shops, ash pits and other such facilities is a major issue both in terms of cost of operation and tax bills.
2) Crews. Firemen are unnecessary on diesel locomotives, but very much necessary on steam engines, even on steam locomotives with stokers. Working rules changes (and one acrimonious strike) caused the idea of crew sizes being reduced to be common by the 1960s. Diesels can easily be operated by one man, which is not the case on a steam locomotive. In the 1950s that's not a problem, but as financial issues come around, it will rapidly become one by the 1960s. On top of that, all the extra shop guys is another substantial issue to deal with.
3) Acceleration. As I mentioned in the ideas thread, diesel-electric locomotives make maximum torque at minimal RPM and get less efficient as speed rises, which is why they are all geared. (This is also a key reason why the most powerful locomotives in America at the time of steam, Union Pacific's gas turbine electrics and Krauss-Maffei's massive ML4000 diesel-hydraulics, had such different fates.) At lower speeds, however, diesels will get a train moving more quickly and easily than a steam locomotive of the same power. This is particularly important for a heavy coal hauler like the Norfolk and Western, particularly before their purchase of the Virginian railroad in 1959, because the N&W had some big grades to tackle on its routes through Virginia and West Virginia. (This is why they bought the Virginian in the first place.) This makes diesels look more efficient on many runs, particularly in the Appalachians. What might work well for you here is to have some diesels on the fleet used for pushing service and pairing them with steam locomotives, or the use of steam turbine-electrics.
4) Modernity. If you want to retain steam, making modern units is an absolute must, not just in performance but also in looks. Rebuilding will only get you so far, which is why I like the idea of steam turbine electrics and the use of electric motors on booster trucks. The NYC Hudsons and the various Pennsylvania railroad experiments (particularly the mighty T1s) styled by Raymond Loewy are a good place to start. Scullin disc drivers (like the NYC Hudsons) would be a must, and reliable booster units should be used across the board. The N&W's massive Jawn Henry steam-turbine electric would be an ideal drag-freight locomotive for heavy-freight trains if it could be made reliable and fitted with dynamic brakes.
5) Traffic. The N&W, like every American railroad, saw freight traffic grow massively in the 1950s and 1960s, and trains got heavier all the time. This is why locomotive power swelled rapidly over the decade, with the most-powerful diesels making power expanding from 2400 hp in 1960 to 3600 hp with the Alco Century 636 and the EMD SD45 in 1966. N&W's merger with the Wabash, Nickel Plate Road and Pittsburgh and West Virginia in 1964 will just about triple the railroad's mileage and massively expand its need for motive power, and butterflying that will likely result in much of the same fate as the Dereco roads in the 1970s as traffic problems hit.
One idea that occured to me is to rebuilt a particularly good base unit (perhaps an ex-NYC Niagara or a Y6b Mallet) with a huge condensing water tender with reversible electric motors on said tender and a pair of high-power generator units on the locomotive itself. Thus, the unit could use the electric motors for better acceleration at slow speeds, overcoming the traction advantage of a diesel, and also allow the locomotive to have dynamic brakes. Air-powered throttles (common on later steam locomotives) could conceivably also be modified to allow steam locomotives to multiple-unit with diesels by using the steam control to control diesel prime movers and the throttle control on the steam engine to control a diesel. (This would be impractical with a diesel controlling a steam engine, though.) This could allow you to have steamers until the 1960s, at which time you will need to rebuild many units for both better efficiency and reduced pollution.
