It might be due to the use of the USB-A ports on the backside of the laptop which are known to have some abnormal power draw, which framework is currently addressing.
They were talking about the usb modules. The back two on either side (the ones closest to the screen) have an issues where the USB 4 Type C to USB A adapter is not fully going to sleep when nothing is plugged in. So ArsTechnica's reviewer most likely had one of the USB A modules plugged into those two USB 4 ports which causes their apparent poor battery life.
Indeed. On both the picture at the top of the article and at the bottom, do we see them having the USB-A expansion card in the slot closest to the screen. If that's how it was configured for the battery test, it would show battery life in a non-optimal configuration.
I believe Framework said they are looking into fixes for this issue, either with firmware patches or if neccesary a hardware revision of the USB A module. Current workaround is just to not put USB A into those two back slots.