Oh I realise that and indeed suggested somewhat similar back in Reply # 4. I was referring more to the R1820 vs R1830 change.
The length difference between the two types is primarily forward of the mounting flange,
and over half of it is the deeper reduction gear housing on the R-1830. The accessory
sections on the rear of the engines aren’t too much different in length and those bits can
be rearranged as necessary.
Realistically the cowling length would only increase by an amount equivalent to the added
cylinder bank. The cowlings of the Cyclone/R-1820 engined DC-3 vs. those of the Pratt &
Whitney Twin-Wasp/R-1830 powered DC-3A is a good guide to the changes.
Ditto the R-1830 vs. R-1820 powered versions of Curtiss P-36/H-75.