After hours of testing, I'm reluctant to say that attributes, such as Int, Str,Mnd, Etc. only decrease burden gain at a capped (25% burden reduction) no matter if your attribute exceeds your pets by 1, 10,25,50,...,999 when using a maneuver, so high attribute sets are not necessary. (As long as your respective attribute is higher than your pet's after using a maneuever you will recieve a 25% burden reduction and a faster burden decay rate, but you will not get any more or less than this.)
As a quick example of what I'm talking about, say you're using Fire Manuevers. Then it has been said when using a "STR" gear set up in conjunction with the fire Manuever it will produce "Good Burden" and reduce burden gain by 25%, as long as your STR attribute (Or any attribute of the respective manuever) is currently higher than your Automaton's STR Attribute (or any attribute of the respective manuever on the automaton).
Most Puppetmasters believe that the more attributes stacked with its respective manuever, the less burden they would gain, but this isn't the case... I've done two simple tests while closely monitoring burden gain.
First Test: Using all the "Optimal" gear for every attribute
STR - Fire Maneuver
INT - Ice Maneuver
AGI - Wind Maneuver
VIT - Earth Maneuver
DEX - Thunder Maneuver
MND - Water Maneuver
CHR - Light Maneuver
MP - Dark Maneuver
In this first example I tried spamming fire manuevers at STR +0, STR+30, and STR+78 on the test server and tested burden gain and overload rate incomparison to no STR gear on. In result, burden rate for STR+0 overloaded at a higher rate than the STR +30 and STR+78, but the STR+30 and STR+78 both gained burden and a 25% reduction and niether cases were better than the other. Burden also decayed at the same rate for STR+30 and STR +78.
In my second test I tried using a Primival Brew and comparing it to the same data collected from above. With STR+999 and spamming Fire Manuevers Burden still decayed at the same rate as STR +30 and STR+78 and also recieved a 25% reduction.
*Keep in mind all of this testing is being used with Stormwaker Frame in order to insure that STR of the master can easily exceed the Automaton*
In conclusion, this proves that creating high STR/MND/AGI/etc. macros for maneuvers is not necessary to still recieve the same reduction in burden and burden decay, and Puppetmaters can get away with not using these type of macros at all granted that most attributes on the master are higher than the automaton to begin with.