Even with 64-bit integers you have 20 decimal digits to play with, and the current implementation only uses 16 (8 digits before the decimal, 8 digits after), so the .01 uBTC precision limit must be a design decision not a datatype limitation.