If you look at the dipstick guide, you'll see that it's not straight. As a result, fluid is always smeared all over the stick. That's why you'll find the reading is unreliable, sometimes different on either side of the stick. The diptube method always works, once you figure out the calibration.

I've got some spare weed eater fuel line that I plan to use for a dip tube. I was thinking I'd make it the same length as the existing dipstick (tip to the point where it contacts the lid) and mark it with some tape there, so that I have a line telling me how far to insert the tube. Then I should be able to mark the tube at the same location as the marks on the dipstick and it give me an accurate reading, correct?

__________________
2010 Mercedes E350 4Matic - 77k miles
1986.5 Nissan D21 V6 4x4 - 276k miles
1987 Mercedes 190D 2.5 Turbo - 3??k miles
1983 Mercedes 300SD turbo parts car
2017 Honda CBR500R - 15k miles
2016 Suzuki GSX S1000 - 9k miles

What I would suggest is take a long length, say three feet. Push it all the way to the bottom of the pan, so that you can always match that level. it should be long enough so that there's no chance of losing it down the hole. Then mark it at the top of the tube guide. Match this mark to the top of the stick, and then transcribe the level marks. Depending on what material you use, use a file or soldering iron to make your marks.

Next time the pan is off, you can verify your marks from below, as the dipstick will be visible at the right forward corner of the transmission.

Excellent suggestion. Thank you very much!

__________________
2010 Mercedes E350 4Matic - 77k miles
1986.5 Nissan D21 V6 4x4 - 276k miles
1987 Mercedes 190D 2.5 Turbo - 3??k miles
1983 Mercedes 300SD turbo parts car
2017 Honda CBR500R - 15k miles
2016 Suzuki GSX S1000 - 9k miles