But you can see if a DL joins the attack, the rating patterns and the actions they do, how often they appear in any animation, etc.
Still yeah, I understand your point, a test vs a much lower avq team will have the advantadge for you so basically you guarantee the attack animations almost all the time, so it may be more profitable to compare attacking players than defenses.
Maybe a more close quality or even a higher team would expose in a better way the defences and the GK in first instance, which is the main player in defense.
But as I mentioned, all is about perception, check the player behaviour and try to feel its limit. If works, works, if do not work, and for some reason the team doesnt play fluidly, maybe there is someone to be sacked.