Effect of Multiple Mask Order (not submasks)

"Changing the mask order has effects."

Please give an example with screenshots, so we're all talking about the same thing. Adobe unfortunately uses the term "mask" ambiguously.

The first photo below shows 2 different local adjustment masks on the same image,

the left circle applies Saturation 0 and the right circle applies Saturation 100. 

So the overlap of those two results in a net zero override (same as saturation 50) to the starting saturation of the imported Raw. The non-overlapping areas show in an uninterfered way, the full override that each mask is imposing.

Here's another test with the adjustments applied by the same two masks reversed (and I also renamed them for clarity) 

Looks as if they are overlapping the same way, giving a net zero override of saturation when combined.

In other words if you apply in effect first + 50 saturation and then -50 saturation, that is the same outcome as applying first -50 and then +50, also the same as doing nothing at all, provided that these saturation adjustments are working non-destructively in their sequence.

Now let's look at how these two will respond to a Basic panel global adjustment of +100 saturation.

First, the version with desaturated area on the left imposed by the lowermost adjustment in the Masks panel.

And just for completeness, with desaturated area on the right imposed by the uppermost adjustment in the Masks panel.

Same result so far as each mask, same result so far as the overlap, both responding the same to global adjustment..

AFAICT from this, the relative ordering within the Masks panel does not seem to matter (mask subcomponents are a different matter). If one mask's desaturation HAD quasi-destructively denied the other mask anything to then increase the saturation of, these tests would have varied in the overlap area.

It also I hope satisfactorily shows the dependent sequence of first local, then global adjustment. Incidentally global Tone Curve then follows after global. If a given area of the photo shows blown highlights under given constant global settings, then a local adjustment (darkening) can in effect retrieve those so they are not blown, providing that this highlight detail is present within the camera file of course. But Tone Curve cannot be used to reverse this blowing-out of the highlight detail. It can only work from what it is given.