Cecil J. Allen has pointed out in a book (The London and North-Eastern Railway, about 1969) that the cylinders and valve chests of the Raven pacifics were manufactured as a single casting, and that after the advantages of long-lap, long-travel valves had been shown after the 1925 exchange between GWR "Castles" and LNER A1s, that similar modifications to them were too costly to modify. Clay and Cliffe (The LNER 4-6-0 Classes) have pointed out the difficulties of fitting a decent-sized ashpan to a large-wheeled, large-boilered 4-6-0. I know the GWR managed it, but they seem to have been the exception. It may account for the indifferent steaming of some of the early express passenger 4-6-0s, as they reached the end of a trip.