Thanks for your further explanation. I think it must be a complex situation. For example, even if a penguinized site escapes from Penguin in the future, its new rankings could depend on whether any competing sites also escape, which the Google Accessible Search has no way to predict. Also, at this point in time Google Accessible Search probably doesn't incorporate any effects that would result from disavowing backlinks. So as you said, it might help to identify a possible problem, but probably shouldn't be used to try to predict what would happen to a site's rankings if the problem is solved.