Whilst I don't have any technical experience regarding steam locomotive performance I have nonetheless read several articles regarding the "enhancements" to steam locomotive performance that took place in the last decade or so of steam operations in South Africa, the WCR "conversion" of R711 and relatively recently the conversions done by the previous owner of Flying Scotsman 4472. In very broad terms South African railways experimented with several of their 25NC class 4-8-4's with particular emphasis on improving the drafting in the firebox and smokebox to improve combustion and thermal efficiency. The outcome of this was sufficient improvement in operating performance to lead them to undertake even more extensive modifications to another 25NC that ultimately became the sole Class 26 member known as the Red Devil. In typical SAR fashion the tests were detailed and thorough and several technical articles were published about the modifications and the resultant improvement in both performance and coal and water consumption.
I was fortunate enough through the late 1980's to ride behind both the modified Class 25's and the Class 26 on some incredibly impressive runs on both passenger trains, chartered steam safaris as they were known and freight services. One of the most exhilarating of these was a run between Kimberley and De Aar in April 1991 behind a twin stack (side by side) modified Class 25 on a 14 car steam safari consist with a number of noted enthusiasts carefully logging speed and running times. The train peaked at over 140 km/hr on a number of occasions and rain for extended distances at 120km/hr.
The Alfred County Narrow Gauge Railway south of Durban also modified its NGG Garrats with a similar range of modifications that achieved enhanced performance and it is my understanding that the people at WCR had quite a lot of contact with the South Africans when they modified R711. I too had the opportunity to ride by R711 unassisted on the Saturday Warrnambool runs and whilst it is purely subjective a chat with the driver on the day plus my own experience from riding behind R711 compared to other R classes was that there was a pronounced improvement in performance. In later years I heard chit chat that the improved performance of R711 resulted in the locomotive having issues with its running gear and the boiler being "forced" that lead to it requiring extra maintenance attention but as I say totally unsubstantiated. Similarly the owner of Flying Scotsman 4472 prior to its recent acquisition and refurbishment by the National Rail Museum also undertook very similar modifications to 4472 and there is a Youtube video of its first run following the modifications with interviews from both the owner, his appointed engineering manager and drivers on the day claiming considerably improved performance.
So whilst none of this has any technical data to back up such claims it is interesting nonetheless that quite work was devoted to trying to improve performance of steam locomotives long after their demise from regular operations and it is a shame in some respects that there wasn't able to be some real technical analysis done with R711 to truly determine if indeed the claims were true.
The difference between the SAR Red Devil (26 class) and the WCR R class is that the improvements to the 26 were properly tested and documented by the Engineer in charge. He subsquently published a book (D. Wardale "The Red Devil and Other Tales from the Age of Steam"), which is very hard going even for a fellow Engineer to read, but contains a wealth of information. We will never know for sure now about the WCR R class. None of the performance calculations I did from trips I went on showed any improvement over a standard R class (and ofc any steam loco can produce much higher power than its official maximum, for short times, if the crew are keen). R 766's last run, its return to Ballarat hauling some WCR diesels, did give some better figures, but I was not on this and was relying on another person's information, and thus I cannot be sure.
It is notoriously difficult to rely on opinions and 'seat of the pants' assessments of loco performance. One very well respected steam man who worked on the 26 in Africa has a number of times told me of fantastic exploits he witnessed with this loco. The trouble is, when you take his figures (grade, speed, tonnage etc) and do some calculations, the Red Devil needed to be producing over 9,000 hp to achieve it!
It was said, in steam days (I have read, I'm not that old!) that the only thing you had to do to improve a locomotive's performance and efficiency was to paint the chimney blue. What this meant was that as soon as the crews thought a loco had been modified, and was being monitored by the technical staff, they would suddenly start doing all the things they should have been, instead of being typical human beings, and taking shortcuts, not caring and being lazy, and equally suddenly that locomotive would start performing much better. The performance of a steam locomotive is more dependant on the crew than anything else (excluding obvious external factors, like bad coal, or severe wear).
I have recently read a theory about the modified exhaust systems such as fitted to the 26 and the WCR Rs that made me think, and I can see some merit in it. The theory is that you can design a better exhaust system for producing maximum power, but that this compromises performance in the normal range of power required for day to day operation. This makes sense, as if you design an exhaust for a certain steam mass flow rate (at max power), this gives different proportions to that if it was designed for lower "every day" power (steam loco rarely use full power, unlike diesels). The exhaust may not be ideal for both. It is a bit like when we used to be able to 'hot up' a car (pre electronics) by adding extractors and a bigger carby (remember those?), yes, we got more power at the top end, but it ran poorly at lower revs, where you usually were in traffic! Maybe this could go some way to explaining why these locomotives rarely live up to the claims made for them?
I understand that the changes to Flying Scotsman were different to the deliberate mods made to locos like the 26 and WCR R. It had gained an A4 boiler because one was available when it was needed, the cylinders were bored out simply because of wear, and the double chimney was a standard feature applied to some A3s later in their lives.