Since I recently went back over the topic of the Somua S35 tank, I felt like I would leave some comment on one part of it that heavily limited its potential: the engine.
The S35 used a 12.7L 60° V8, delivering 190hp at 2000rpm (normal regime) and 200hp at 2200-300rpm (maximum regime):
The origin of this engine was the Janvier Sabin & Cie automobile company which designed and built some cars but mostly engines for other companies. It closed down in 1928. It is now believed that the Somua did not contract this engine, but actually just bought the drawings of an unbuilt design for use in the S35 as no other Somua engine was suitable. The 20's origin of this design, and its likely intended use in big cars or trucks as a big block V8 probably explain some of its features. The engine apparently uses reverse flow heads which provide good performance at low rpm, but is heavily limited at higher rpms. The updraft carburettors it uses are common on truck engines of the era, but not on performance engines. Finally it appears to use a bathtub-type combustion chamber which is also not ideal at higher rpms.
The end result is that, compared to high performance engines seen in contemporary tanks, it somewhat suffers. The Hispano Suiza J12, used since 1931 and seen in the ARL V 39 assault gun and the SEAM G1P battle tank, could deliver 220hp in its early 9.4L version (a 10% output increase on nearly 75% of the displacement), or 250hp in its 11.3L version (10% extra power compared to the bored-out 13.7L engine in the S40, with 80% of the displacement). Hispano-Suiza even was working on a 280-300hp version of that engine for production ARL V 39s as of 1940.
Meanwhile, the contemporary Maybach HL 108 could still achieve 250hp on 11L, and the HL 120 could achieve 265-300hp at 12L. Czech tanks like the Skoda T-21 and ST vz.39 had similar 12L-class engines with more power than the Somua.
The SOMUA was thus missing anywhere from a 10% up to a 50% increase in power output within the same volume in 1940. This would have heavily limited the growth potential in terms of weight and mobility, as bigger turrets could already increase weight by 2 to 3 tonnes alone.
There is a silver lining to this however: if the drawings for the engine indeed dated from the late 1920s, it could have been a perfectly suitable powerplant for Renault's D1, D2 or even possibly AMC 35 tanks from the early 30s, which used weaker and sometimes even bulkier inline 4s or inline 6 engines. This was demonstrated in the S35 which could still boast more power than the D2 at the same weight, same dimensions, same armament and same armor.