Yes that's true, if the water changes obviously your whole calibration is out.

But how about this...couldn't one make 4KH reference solution or even any KH, doesn't matter; and then test it to find out the exact atmospheric equilibrium in your location, and use that method to determine the amount of co2 after offgassing a sample of tank water? You would need a pH test kit that is fairly accurate, to .1 probably, for it to be useful. But if you use a KH that's relatively easy to see the difference on the chart, you could be pretty accurate right? At least probably within 5 - 10ppm of the real amount of co2 in your tank? I'm not saying that this supercedes drop checkers, but for people who don't have one.

Or how about this as an alternative to figure out if your water "works" for the chart... take distilled water and add enough KH to bring it up to your tank's KH, then test pH on both your tank and this water. It should be very simple to compare two colors and see if there is any difference at all one way or another. If you can't tell the difference that way, you won't be able to tell on the chart either.