Not sure if referring to my GTX3067 result, or typo, but nonetheless also found the results of 0.81 v 0.96 very interesting.
Below quote from the original page in particular:
The .96 A/R only added lag and if you reference the boost spool comparison below, we also weren't able to see as much boost with the duty cycle all-in. This tells us that we were at a limitation of the compressor wheel flow itself and not because of back pressure.
Not always beneficial to go larger A/R it seems.
Anyways, sorry to digress from subject of thread. Thanks Lithium for posting above info.