Also looking at the civs in the test, besides the tested ones I am only seeing Murica, England, Korea, Netherlands who typically are middle of the pack at best and usually punching bags, with many variables on how they perform - how are the start positions in the pictured scenario? Korea especially tends to do badly unless it goes Tradition and even then it can fail if it's a target, and Netherlands pretty much never do well - probably thanks to an UA the AI can't utilise properly and a pretty crappy, late UI. Even if the other civs were of the "always do well" type like Songhai, I still wouldn't really be convinced because I can't see the terrain. I can remember one game many versions ago where Venice somehow did pretty well because it built almost all the wonders and the AIs around it for some reason didn't conquer him (so I did it instead) so he got a pretty decent lead before his death.
The map you present also has a very low land percentage... Not much place to settle cities. Better than the one I remember you showing as testing on last time (each civ had place for 2-3 cities), but it's still small.
I know you do many more tests and probably some meet my suggestion where they probably do well anyway, but testing those changed civs in at least one game with better, more consistent AIs like Songhai, Russia, Celts, Carthage, Portugal, Maya would be better - if it didn't occur already.