Wouldn't this lead to bigger error accumulation? At least, my gut feeling says "less information per delta = larger error accumulation", and on top of that we h

Wouldn't this lead to bigger error accumulation? At least, my gut feeling says "less information per delta = larger error accumulation", and on top of that we have more deltas/second