Assuming that the melamine is in excess and that the CYA is always the limiting reagent, then yes, any tube should be equivalent (provided they are based on a 50/50 mixture). I believe that's the case, as the CYA levels are unknown before you perform the test, so there must be sufficient melamine to bind up all the CYA and create sufficient turbidity to get a measurement.
I theory, you could make your own tube. Just put something contrasty at the bottom of a tube of some sort, and use solutions where you know the CYA (i.e. prepare some CYA solutions of known concentration) to calibrate the tube, marking off where you cannot see the markings at each level of CYA.
Of course, it's easier to spend a few bucks and just buy one, and that's what I did.
The scale is (probably) logarithmic. If you notice, the spacing between 20 and 30 is not the same as between 30 and 40, or 40 and 50, etc. I'd have to think more to figure out exactly what equation it is, but I'd guess it's a logarithmic scale. However, the exact distances, while the ratios should be consistent, will vary based on the contrast of the markings at the bottom.
Personally, I hate this test. I do not like figuring out at just what point I can or cannot see the stupid black dot. If I had a turbidity reading machine, that would be one thing, but not many of us have those.