I too am not sure about the Leader - I admire the fact that Bulleid was prepared to challenge conventional thinking - and was recruited by the SR to modernise their non electric part of the railway, but the Leader was always going to be a nightmare for the fireman and the technology of power bogies/sleeve valves and chain drives in the average running shed in the 50's was also waiting to cause problems.
Yes, the "average running shed in the 50's" was pretty hostile to anything bar a standard Rocket-on-steroids steam loco; it could be argued that it's less of a surprise that Leader failed than that diesels didn't. As it was, it was that they weren't going to be allowed to that made the difference...
But it wasn't a fundamentally bad design, it was just not well suited to the conditions of the time and the lack of resources to debug it, especially when trying to do it on a busy full-size railway already struggling with post-war dilapidations. A lot of experimental designs ended up being canned because the later stages of development necessarily involved plonking a half-completed thing down on an operating railway and requiring that it performed no less reliably than the tried and tested designs already running. The kind of situation we're talking about, where something as major as a complete locomotive is being built just because a bunch of people think it's fun to do, is considerably less fraught (or can be, at least) and so it is more possible to take it to the point where success or failure speaks for the design itself, and not the dissatisfaction of a management with plenty of more pressing concerns to worry about.
Moreover, since it was experimental, there is no justification for regarding those features which the original experiment showed to be dubious as being set in stone; on the contrary, improving their design is fully in the spirit of the original project. Optimising the design of sleeve valves is a whole sight easier when you can accurately predict thermal distortions using finite element modelling with free software on any old PC, and when technologies such as making parts out of one substance for the bulk material and another substance for coating wear surfaces are so much better developed. And I think using oil firing for the sake of the poor old fireman goes without saying!