A couple of points about the engineering if I may. The idea of the chains ‘elongating’ rather than stretching is confused by the terminology. It is a particular feature of Morse chain, a type of silent chain. It is correct that wear in the pins and links will cause the pitch, that is the distance between the centres of pins in a link pair to increase. This is true of a roller chain, think bicycles etc. The difference, is that as the chain wears and the links move further apart, they follow the teeth of the gears, so that although there is now a gap at the bottom of the teeth on the gear wheel, the teeth on the inside of the links are still meshing nicely with the gear teeth. The chain meshes with the gear in the way that gears mesh together. Motorcycle engines have a version of these chains, commonly driving the camshaft(s) and sometimes the primary gearbox shaft. A roller chain just gets sloppy and loose when it wears and meshes badly. That’s why a worn bike chain is noisy. Look at a bike with a worn chain if you can, to see what I mean. I can’t find any pics as evidence yet, sorry. As for cardan shafts, the drive would have needed two shafts, each with two universal joints, assuming they followed the route of the chains. I can’t see another way of connecting them to the crank and valve gear shaft other than with bevel gears. The arrangement for the crank would be like a live axle on a lorry. The idler shaft would need two sets of bevel gear to go round the corner. Were the bearings in the original design rolling element type or plain? The more expensive and in wartime scarce roller bearings would mean that the lost motion remains constant for a long time and can be allowed for in the design. I’d love to see the original layout with shafts, if it exists. Lastly, the wheels. They are not a boxpok wheel as is frequently stated. They are a wobbly web wheel, a type of disc wheel. I have seen then called a wavy wheel as well. GIYF here.