Let me start with another quick example. The beloved Eureka (pauses for a moment lost in day-dreamland). That engine has had a rough service life. All new wood cab built by Dan, new lead truck (1880s design), new side rods (probably replaced same time as lead truck), and new reverse motion parts (more 1880s design). That's a fair portion of the engine that's not original, but I don't care, in fact nobody cares. It's a beautiful OPERATING engine. The only 99% original engine I know of (there may be more) is the Eureka's sister, Sonoma, which is mothballed at the Cal State RR Museum. Given the choice I would rather spend time looking at the Eureka.
The next question in mind is about preserving equipment for future generations. An engine kept in proper working order and maintained (replace parts as needed) will last effectively forever. One on display is literally rotting away. All metals are slowly degrading back to their most primitive form. The most pure steel made slowly pulls carbon atoms back out of the air as it progresses back to a state of raw iron. Even the metals in Stainless Steel are slowly pulling themselves apart. None of it will last forever. If you look closely at the Sonoma's iron drive wheel tires you'll find flat spots from sitting in one spot too long. The wheel center castings are probably warped as well. How much longer will it last?
Onto the topic of restoration. It's obvious by now that I like operating engines versus dead hulks so I won't say any more about this.
Onto the topic of replicas. I've spent many years studying the designs of Baldwin locomotives (mainly 3' narrow gauge of the 1874-1900 period) and came to a few decisions. I would never build a 100% replica because as it's a waste of time/money and too much of a headache. I would redesign my replica to use roller bearings on all axles without a second though. I would also hide roller bearings in the side rods and any other rotational / rocking motion pieces, including the valve gear assembly. It would just make a far more efficient engine and they can be hidden to the point it doesn't effect the physical appearance. I would also use mostly Lost Foam castings. Green Sand is too expensive thanks to environmental regulations and true pattern making in a lost art. Lost Foam costs MUCH less and is flexible enough to handle the most complex castings on the engines (the cylinder block) without any problems. Of course if you get within 5 feet of the engine you'll see it's not made of "proper" castings, but that can be covered up with a bit of work so that no one would know by looking at it. I would also have a whole new boiler designed for higher efficiency, while only keeping the outward shape. Basically, if it can't be seen I would modernize the design as much as I could. I would also use as many space-age metals and alloys as I could rather that plain iron, so long at it looks right.
That would make an engine of lower cost to build that a 100% replica and it would be much more fuel efficient. Both of which are good things to have in this day and age.
But that's just me & my opinions.
Cheers
Curtis S. Ferrington