It's a matter of accuracy.
Going from 10N to 0.2N means diluting 50 times.
Say you want 100 mL of the 0.2N solution, and you do it in one go. Then you'd have to measure out 2 mL of the 10N solution, and how accurate can you do that? With the graduated pipette I have the accuracy is +/- 0.02 mL which translates to 1% inaccurate.
If you first measure out 10 mL (with the same +/- 0.02 mL) there is a 0.2% inaccuracy in the making of 100 mL of a 1N solution.
You then measure 20 mL (+/-0.02 mL) of the 1N to make up the final solution, with an inaccuracy of 0.1%.
In total, the inaccuracy cannot be larger than 0.3%. In practice (statistically), the inaccuracy will be less, closer to 0.25%.