Well look at it this way. Unless you invest in an expensive scale that can read down to the 0.01g, you have an error range of about 100mg (like I said before, 0.6 reading could be anywhere from 0.55 to 0.64). This would make it very inaccurate to dose at once.
If you took lets say a week's dose (say 600mg/day, 4.2g) and measured that out, there would be a 100mg discrepancy.
By dividing this powder into 7 roughly equivalent piles, youd narrow down the error (assuming your eye is good). Now the daily dosing error would be something like 14-15mg (assuming once again good eye measurement).
Splitting this daily dose into 3 doses would then reduce the dose error to 5mg. 200mg with a +- of 5mg is alot better than 200mg with +- of 100mg.
Even if you didnt do the week into day into dose, just doing the day into dose would reduce the error in daily dosing.
You'd still get that +- 100mg difference, but here it would be spread out among 3 or 4 doses instead of having each dose being +- 100mg error.
See where I'm getting with this? It would probably be much easier btw to split the 600 into 3 or 4 as opposed to splitting the weekly dose into 7 adn then into 3 or 4.
Or you could invest in a 0.01g accuracy scale.