Some details of the Roger-Smith family:
Company records show an Edward Mervyn Roger-Smith, b. 1941, as of Priory Farm, Alton, Hants.; electoral roll records show Edward M. Roger-Smith as in the same household as Alexandra R. and Henrietta E., alongside other children.
Given the 1988 birth record for Henrietta Elizabeth Roger-Smith, mother "Chouffot" (alongside those of her elder brother and three sisters, corroborating the electoral roll etc), the 1971 marriage of Edward "H." Roger-Smith and Sandra R. Chouffot would appear to be the correct one ("Sandra" of course being a commonly-encountered version of "Alexandra", if odd to see on a marriage record; perhaps the former is her actual name and the latter a variant she prefers).
No birth record for an Edward M. Roger-Smith (nor Edward M. R. Smith, or even a likely Edward M. Smith). A family of the name appears in Ruvigny's Plantagenet Roll of the Blood Royal, following the descent of Dorothy Eugenia Woodd who m. 1900, Hugh Roubiliac Roger-Smith, M.D., M.R.C.S., L.R.C.P.; their two sons were Raymond and Basil; the latter's details can be found here-
https://www.chch.ox.ac.uk/fallen-alumni/captain-basil-hugh-roger-smith ; corroboration of the birth of his son discounts Edward M. Roger-Smith. Raymond seemed a likely option, but his death in 1942 in Natal, South Africa and corresponding (very limited) probate record indicated him to have been a bachelor. I had thought that the possibility of a South African birth for E. M. Roger-Smith would neatly explain matters, but clearly not!
Per records, Hugh Roubiliac Roger-Smith was b. 1867 at Lewisham; the birth record gives only "Smith", the 1900 marriage record giving both "Hugh Roubiliac Roger-Smith" and "Hugh Roubiliac R. Smith". The 1891 census has Hugh R. Smith, medical student, b. Kent, in the household of "Ravens Croft" (sic?) E. Smith (1860-), architect, and Edith Smith, along with a visitor and three servants.
A search of architects gives Ravenscroft Elsey Smith b. 1859 d. 1930 as s. of Prof. Thomas Roger Smith (1830-1903), and in business with his father as well as Professor of Architecture at King's College London and University College London.
I have to leave this here for now, but can try and pin down E. M. Roger-Smith later. On the current evidence, I'd imagine he descends from Thomas Roger Smith, if possibly of a different line from that of R. E. Smith.