In Real Battles As apposed to blank terrain firing test simulations, The main difference being the number of shots fired at reduced accurracy due to terrain and movement very few shots at front to front and the inclusion of a huge number of variables from poor identification to distraction by infantry and other anti tank fire. I am trying to get across the limited value of range testing when figuring out real (In game) combat effectivness. In addition I have played in excess of 20 different table top simulation systems from the basic to highly complex. All these come from authors who have different views and opinions that "colour" results. A game engine PC based or tabletop will always be innacurate to a point as a fuction of the limited variables and how they are applied in each system. An example is the application of angle modifiers to armour thickness. some systems take an average angle modified armour, others attempt to sub divide it, others yet apply the angle to the shell penetration factor. These will all give slightly different results under a specific same situation.
And no I haven't fought panthers vs shermans in real life, Thankfully I never had to and no one will in the future.