Implementation counts for a lot as we have seen in other threads
As an observation the 252 should perform slightly better than the 122 in the above chart.
Also the 502 test had a measurement set up issue as discussed in the thread so does perform better than indicated.
The issue may have been discussed, but was it resolved? In post #87 on January 16th, in this link:
https://www.audiosciencereview.com/...02-stereo-amplifier-measurements.10909/page-5
A drawing was promised for the next day after the post to help clarify the issue, but continued reading on that thread, and a quick review now, finds that you have yet to post the promised drawing.
If the drawing is in the tread please provide the post number, otherwise the drawing would still be appreciated since the drawing covers some interesting questions.