For an alphabet, with each letter we have a minimum number of dots/ pixels required to represent it.
For convenience one takes such a value for the whole alphabet but actually in addition to this it varies for each letter rg number of dots required to denote an 'I' is lesser than 'X' .
Comparing letters of different alphabets 'X' certainly has lesser requirement of dots to be recognisable than 'ध' ( 'gha' ) in devnagari.
Given the angular resolution of the pixel as subtended to the eye remains the same... The metric is agnostic of letter size.
**Question: Using which alphabet can Hamlet / Dhammapada be transliterated so that the book has least number of dots/pixels ( and yet is readable) **