When you draw fluid from the dipstick tube, you're really only drawing fluid from the sump. This leaves a lot of fluid in the torque converter as well as fluid passages. And this old fluid will dilute and mix with the new fluid.

The best way is the exchange method through the cooling tubes, but it also has the problem with mixing old with new. You have to add a little at a time, each time the new fluid being mixed with the old, so you have to add/extract a lot of fluid to do a thorough flush.

I've worked this out mathematically (using Excel to generate iterations) with the 240's AW70 as a model. It's not the same capacity, but it can serve as an example to you:
We're assuming a 7.4 qt total capacity (including torque converter), and exchanging 2 qts at a time (each iteration).
after 6 quarts (5 iterations), 47% of the fluid is new fluid.
after 10 qts, 72% of the fluid is new.
after 16 qts, 89%
after 20 qts, 94%
after 26 qts, 98%
after 30 qts, 99%
As you can see, this is an asymptotic approach to 100% -- you reach 100%, to 4 decimal places, only after 52 quarts!

You've got a larger capacity, so you'll need more quarts to achieve these concentrations of new fluid.