Section IX QW-408.2 recommends the use of AWS classification of SFA-5.32 to specify the shielding gas composition.
So...in the case of this post, if the WPS specify SG-A of SFA-5.32 , then gas purity change from 99.997% to 99.999% is not considered a esstential variable and do not require requalification of the WPS.
However, since QW-408.2 says "may", the use of SFA-5.32 classifications is not mandaory. The way the WPS specifies is OK. I assume that the qualification test was done with a specified 99.996% purity Ar~
Note that the backing gas is another esstential variable as dictated by QW-408.9 for groove welds of certain materials~ By specifying "high purity argon min. 99.996% for both the torch and backing gas", the WPS has specifed two variables of Section IX. Whether QW-408.9 is esstential or not depends on the welds and materials.