I suspect you will find a lot less of these "bad things" than anticipated. That is why the model should actually be freely available rather than restricted based on pre-conceived notions that will, I am sure, prove inaccurate.