I’m also failing to see how digit efficiency would be important in DNA. In fact, it seems that a high base system would be more efficient. If you had 80 nucleobases instead of 4, each base pair would contain far more information
Strictly speaking, you could encode more error-correction bits into a higher nucleobase count but that'd pretty much require intelligent design, and wouldn't necessarily be viable for microorganisms to have that many proteins handling all the metabolic scaffolding.
The efficiency comes from the ratio of the alphabet to the number of character places needed to express them. Otherwise why not base a million? Or a billion?
This ratio is what leads to base e being the theoretical maximum.
