While I’m not a “nitty-gritty” aerodynamicist I think its accurate to say that because airplanes operate in the real world, they are subject to nicks, dings, warps, stretches, wear ‘n’ tear, and the like … meaning that small things can, and often do, generate their own individual inputs that may not be as apparent as some other larger anomalies. Also, any cockpit video is at least somewhat incomplete unless it includes visuals on all 3 of the control “appliers” in the cockpit – column (pitch), wheel (bank), and rudder pedal (yaw) – including some sort of readable scale - and it’s a lot harder to get good comparative video of the rudder pedals simply due to their location.
Two things about the "data" upon which the "aero-model" is based…
1) airplane data should be the primary contributor to the aero-programming (and it is usually supplemented with a good understanding - or reference materials - as to what was intended to be the aerodynamic result) and…
2) a well-qualified and experienced pilot’s input regarding nuances can be invaluable.
And the final recognition should always be that a simulator (the best kind of simulator, included) is not the airplane. Any training conducted in that simulator should be conducted by someone with considerable knowledge of that airplane AND that (specific) simulator - with particular emphasis on what THAT simulator does, does well, and does not-so-well. Or, said differently, un-supervised "practice" or other use of the simulator, should be very heavily scrutinized with a full understanding of the influence that can be generated by that very powerful training tool (i.e., the simulator).
Getting someone to the point of being able to fly the simulator well is NOT and should NOT BE, the goal of training. We are teaching pilots to fly the AIRPLANE ... and the simulator is a tool to help do that ... it should be acknowledged as being a viable, incredibly capable tool, and much more so when used to its full capability, but, we MUST recognize that it can generate negative issues when it is depended upon to impart the knowledge instead of assisting the instructor in doing that job.